fix(snowflake): migrate to native endpoints with auto-routing for Claude models - #29964
Merged
Sameerlite merged 23 commits intoJun 12, 2026
Conversation
Replaces the legacy /api/v2/cortex/inference:complete endpoint with the native OpenAI-compatible /api/v2/cortex/v1/chat/completions endpoint, fixing error 390142 (Incoming request does not contain a valid payload) when using model: snowflake/<model> in LiteLLM proxy. Changes: - litellm/llms/snowflake/chat/transformation.py: route to native /cortex/v1/chat/completions, remove Snowflake-specific tool_spec payload transformation, remove content_list response handling, add stream to supported params - litellm/llms/snowflake/anthropic/transformation.py (new): SnowflakeCortexAnthropicConfig routes Claude models to /cortex/v1/messages with anthropic-version header and Anthropic->OpenAI response transform - tests: 29 unit tests covering URL routing, auth headers, payload format, and response parsing
|
|
Contributor
Greptile SummaryThis PR migrates the Snowflake Cortex integration from the legacy
Confidence Score: 5/5The change is self-contained to the Snowflake provider directory and its tests; it fixes a confirmed customer-facing breakage and is backed by live test results across 12 models. The routing logic, message transformation, tool coalescing, and streaming handler are all correctly implemented and thoroughly tested with mocks. Two minor edge cases exist: message_stop hardcodes finish_reason=stop (currently harmless because CustomStreamWrapper discards it), and a user text message immediately following tool-result messages produces consecutive user turns (Snowflake accepts this). Neither affects the reported bug fix or current production traffic. No files require special attention.
|
| Filename | Overview |
|---|---|
| litellm/llms/snowflake/chat/transformation.py | Major rewrite: adds auto-routing to native Snowflake Cortex endpoints (/messages for Claude, /chat/completions for others), with proper Anthropic format conversion including tool_use/tool_result coalescing and streaming handler for both SSE formats. |
| tests/test_litellm/llms/snowflake/chat/test_snowflake_chat_transformation.py | Existing tests updated to reflect the new native endpoint behavior: tool/tool_choice transformations now test pass-through for OpenAI format (non-Claude) instead of the removed Snowflake tool_spec format; endpoint URL assertion updated from inference:complete to cortex/v1/chat/completions. |
| tests/test_litellm/llms/snowflake/test_snowflake_native_endpoints.py | New comprehensive test file (718 lines) covering URL routing, auth headers, request/response transformation for both Claude and non-Claude paths, multi-turn tool conversations, parallel tool result coalescing, and streaming format detection — all using mocks, no real network calls. |
Reviews (15): Last reviewed commit: "fix(snowflake): handle message_start eve..." | Re-trigger Greptile
![greptile-apps[bot]](https://avatars.githubusercontent.com/in/867647?s=60&v=4){kind=small}
Codecov Report❌ Patch coverage is
📢 Thoughts on this report? Let us know! |
…ormat conversion - _extract_system_and_messages now preserves tool_calls from assistant messages and converts them to Anthropic tool_use content blocks - tool role messages are converted to user role with tool_result content blocks (as required by Anthropic Messages API) - Added _transform_tools_to_anthropic() to convert OpenAI tool format (type/function/parameters) to Anthropic format (name/input_schema) - Added comprehensive tests for multi-turn tool conversations Addresses review feedback on PR BerriAI#29964
Contributor
Author
Contributor
|
Thanks for the Snowflake native-endpoints work! Two things to get this review-ready:
Once those are in we'll take another look — thank you! |
…dpoint compatibility
Contributor
Comment on lines
+63
to
+81
| Usage in litellm: | ||
| import litellm | ||
| response = litellm.completion( | ||
| model="snowflake/claude-sonnet-4-5", | ||
| messages=[{"role": "user", "content": "Hello"}], | ||
| api_key="pat/<your-pat>", | ||
| api_base="https://<account>.snowflakecomputing.com", | ||
| custom_llm_provider="snowflake-anthropic", # routes to this config | ||
| ) | ||
| """ | ||
|
|
||
| def get_supported_openai_params(self, model: str) -> List[str]: | ||
| return [ | ||
| "temperature", | ||
| "max_tokens", | ||
| "top_p", | ||
| "stream", | ||
| "tools", | ||
| "tool_choice", |
Contributor
There was a problem hiding this comment.
SnowflakeCortexAnthropicConfig is not wired into litellm's routing — custom_llm_provider="snowflake-anthropic" raises an unsupported-provider error
The class docstring says users should call litellm.completion(..., custom_llm_provider="snowflake-anthropic"), but "snowflake-anthropic" does not exist anywhere in litellm's routing layer:
LlmProvidersenum (litellm/types/utils.pyline 3344) only hasSNOWFLAKE = "snowflake"— noSNOWFLAKE_ANTHROPICentry.main.pyhas noelif custom_llm_provider == "snowflake-anthropic":branch (the snowflake branch is at line 4280).utils.py's config map (line 8323) only mapsLlmProviders.SNOWFLAKE → SnowflakeConfig._lazy_imports_registry.pyand__init__.pydo not import or export this class._is_claude_modelis used only in tests, not in any routing decision.
All tests for this class instantiate it directly, bypassing the routing layer, so they pass even though end-to-end dispatch is broken. To make this class reachable, at minimum LlmProviders, main.py, utils.py, and _lazy_imports_registry.py need a "snowflake-anthropic" entry, and __init__.py needs an import.
…to-routing - Remove separate SnowflakeCortexAnthropicConfig and anthropic/ directory - SnowflakeConfig now auto-routes based on model name: - Claude models → /messages endpoint (Anthropic format) - All others → /chat/completions endpoint (OpenAI format) - No new provider needed (stays as SNOWFLAKE = 'snowflake') - Tool message transformation for Claude: tool_calls → tool_use blocks, tool role → user with tool_result - OpenAI → Anthropic tool format conversion (parameters → input_schema) - Addresses Greptile feedback about unwired SnowflakeCortexAnthropicConfig
…cated max_tokens on this endpoint)
sfc-gh-nashukla
changed the title
Contributor
Author
…n_tokens in Anthropic path
Contributor
Author
…odel prefix on OpenAI path
Contributor
Author
![veria-ai[bot]](https://avatars.githubusercontent.com/in/1723776?s=60&v=4){kind=small}
Contributor
PR overviewAll previously flagged issues have been addressed. No open security concerns remain on this pull request. Security reviewNo open security issues remain on this pull request. Fixed/addressed: 1 · PR risk: 0/10 |
Contributor
Author
Contributor
Author
|
Re: This is intentional and confirmed via live testing. Snowflake's Tested against 8 non-Claude models (llama3.1-70b, llama3.1-8b, mistral-large2, deepseek-r1, llama4-maverick, snowflake-llama-3.3-70b, openai-gpt-4.1, openai-gpt-5) — all succeed with The PR description has been updated to reflect this. The code is correct. Re: |
Contributor
Author
Contributor
Author
|
@greptileai Please re-evaluate this point. The code is correct. Evidence from live testing against Snowflake production: Snowflake's endpoint follows the newer OpenAI convention where Regarding the endpoint migration: the legacy |
Contributor
Author
Contributor
Author
Contributor
|
Want your agent to iterate on Greptile's feedback? Try greploops. |
…ssage for Anthropic
Contributor
Author
Contributor
Author
1 similar comment
Contributor
Author
Contributor
Author
|
@Sameerlite I have addressed all the feedback — Greptile is now 5/5. Here are the screenshots for your reference: Customer bug reproduction — shows "unknown role tool" on v1.88.0 and the fix transforming messages correctly 
|
Sameerlite
changed the base branch from
litellm_internal_staging
to
litellm_oss_staging_120626
June 12, 2026 11:27
Sameerlite
merged commit Jun 12, 2026
489531a
into
BerriAI:litellm_oss_staging_120626
73 checks passed
mateo-berri
added a commit
that referenced
this pull request
Jun 12, 2026
* feat(bedrock): add bedrock mantle gemma 4 models (#30264) * feat(bedrock): add bedrock mantle gemma 4 models * test(bedrock): harden mantle local cost fixture * feat(responses): enable the responses API for the Tensormesh provider (#30209) * feat(responses): enable the responses API for the Tensormesh provider * Update litellm/llms/openai_like/providers.json Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> --------- Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> * fix(langfuse_otel): mark LLM spans as generations (#30250) * fix(bedrock): stop stream_chunk_size leaking into invoke request bodies (#30240) stream_chunk_size is a LiteLLM-internal knob for re-chunking the HTTP response stream. The invoke transformations splat optional_params into the provider request body without dropping it, and Bedrock rejects unknown fields, so any bedrock/invoke request that sets the parameter fails with ValidationException: stream_chunk_size: Extra inputs are not permitted. Drop it in the invoke dispatcher (covers cohere, titan, mistral, meta, ai21) and in the Claude messages-format request builder (the route used for bedrock/invoke Anthropic models) * fix(bedrock): stop buffering streamed tool-call argument deltas (#30231) * fix(bedrock): stop buffering streamed tool-call argument deltas Two issues made Bedrock tool-use streaming arrive as a single end-of-stream burst through LiteLLM while plain text streamed fine. First, the anthropic-beta allowlist mapped fine-grained-tool-streaming-2025-05-14 to null for bedrock and bedrock_converse, so the header was silently stripped. Without that beta, Anthropic models on Bedrock buffer tool input server-side and emit all toolUse.input deltas at once (verified against converse-stream and invoke-with-response-stream directly). Bedrock accepts the beta via additionalModelRequestFields.anthropic_beta, so it is now forwarded. Second, the streaming reads re-chunked the AWS event stream with iter_bytes(chunk_size=1024). httpx's ByteChunker only releases full 1024-byte blocks, so the small early events (messageStart, contentBlockStart, first deltas) sat in the buffer until enough bytes accumulated, pushing time-to-first-byte from ~1.4s to ~8.5s on buffered tool-use streams. The default is now no re-chunking; an explicit stream_chunk_size is still honored. * test(bedrock): cover explicit stream_chunk_size on sync invoke path * test(bedrock): cover stream_chunk_size plumbing through converse completion * test(bedrock): cover stream_chunk_size default in legacy BedrockLLM streaming * test(bedrock): merge converse handler tests into existing mapped test file pytest imports test modules by basename in non-package test dirs, so the new tests/test_litellm/llms/bedrock/chat/test_converse_handler.py collided with the pre-existing tests/test_litellm/llms/chat/test_converse_handler.py and broke collection in CI. Move the new tests into the existing file * feat(otel): emit v2 cost breakdown + stamp tracer scope version (#30156) Read the StandardLoggingPayload cost_breakdown into a typed LLMCost on LLMCallSpanData and emit each component under litellm.cost.* (absent components omitted, so spans stay sparse). Stamp litellm.__version__ as the instrumentation scope version so every v2 span carries a deterministic scope.version. Tests under tests/test_litellm/integrations/otel/. * fix(proxy): cancel in-flight upstream LLM request on client disconnect (opt-in) (#30223) * fix(proxy): cancel in-flight upstream LLM request on client disconnect (opt-in) On the non-streaming path, base_process_llm_request awaited the LLM call with no disconnect monitoring; when the HTTP client went away the upstream request kept running until completion or request_timeout (6000s default), holding a backend slot (e.g. a vLLM GPU slot) for output nobody would read Add an opt-in general_settings.cancel_on_disconnect flag, default off, so the default code path is unchanged. When enabled, a receive-based watcher task observes http.disconnect and cancels the asyncio.gather driving the upstream call. The resulting CancelledError is converted to HTTPException 499 only when the disconnect event is set, so server-initiated cancellations still propagate as-is. The 499 then flows through _handle_llm_api_exception like any other failure, meaning post_call_failure_hook still releases max_parallel_requests slots and fires spend and alerting callbacks; it is logged at info level instead of a full traceback Also removes the dead check_request_disconnection helper in proxy_server.py (zero call sites) along with its behavior-pin tests Builds on the receive-based design from #25776 Addresses #13774. Re-fixes #22805 (regressed after the #14295 revert) Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> * fix(proxy): scope 499 quiet logging to disconnects and harden watcher Address the two P2 findings from the Greptile review on #30223. The info-level logging in _log_llm_api_exception now applies only to the disconnect-specific HTTPException (status 499 plus the shared _CLIENT_DISCONNECT_DETAIL message), so any other 499 raised by hooks or guardrails keeps its full traceback. The disconnect watcher now catches exceptions from request.receive() (e.g. a transport reset) and logs a warning instead of dying silently, making the degradation to no-op visible; a test pins that the LLM call is not cancelled in that case --------- Co-authored-by: kursad <kursad.lacin@brado.net> Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> * fix(bedrock): grant aws-external-anthropic:* in OIDC session policy for claude_platform (#30200) (#30205) The inline STS session policy passed to assume_role_with_web_identity acts as an IAM PERMISSION CEILING — effective permissions are the intersection of the role's identity policies and this policy. Any action not listed is silently denied even when the IAM role grants it. #27678 added the bedrock/claude_platform/<model> route but its service-side action namespace is aws-external-anthropic:*, not bedrock:*. Without a matching statement here, every claude_platform request via OIDC (GCP federation, EKS Pod Identity webhook, etc.) 403s with 'no session policy allows the aws-external-anthropic:CreateInference action' — even with a fully permissive identity policy. Add a second ClaudePlatformLiteLLM statement covering CreateInference, CreateBatchInference, CancelBatchInference, DeleteBatchInference, CountTokens, Get*, List*. Keep aws:SecureTransport=true parity with the bedrock statement. Static creds + IRSA flow through different code paths and are not affected. Fixes #30200 * fix(proxy): set Retry-After header on RouterRateLimitError 429 responses (#30098) * Set Retry-After header on RouterRateLimitError responses When all deployments for a model are in cooldown, the proxy returns a 429 whose cooldown timing is only available by parsing the error message string. RouterRateLimitError already carries cooldown_time, so expose it as a standard retry-after header in _handle_llm_api_exception. The value is rounded up so clients never retry before the cooldown window ends. Fixes #27823. * Set Retry-After after response-headers hook so cooldown wins The cooldown-derived retry-after was assigned before the post_call_response_headers_hook merge, so a callback returning a retry-after key (including a stale or empty value) silently clobbered it. Move the RouterRateLimitError block after the callback merge so the cooldown value is authoritative for this error type. * fix(router): route aspeech through async_function_with_fallbacks (#30104) * fix(router): route aspeech through async_function_with_fallbacks Router.aspeech selected a deployment and awaited litellm.aspeech directly, so TTS requests got no retry on failure and no failover to backup deployments; the except block only fired an exception alert and re-raised. Every other router endpoint (acompletion, aembedding, atranscription, arerank) already delegates to async_function_with_fallbacks Mirror the atranscription pattern: move deployment selection and the litellm.aspeech call into a private _aspeech method, then have the public aspeech set kwargs["original_function"] = self._aspeech and await self.async_function_with_fallbacks(**kwargs). _aspeech also picks up the shared _get_async_openai_model_client helper and the same total/success/fail call accounting the sibling endpoints use Fixes #27778. * fix(router): apply deployment kwargs and rpm semaphore in _aspeech Bring _aspeech fully in line with _atranscription: call _update_kwargs_with_deployment so deployment metadata, model_info, timeout, and default litellm params flow into the request, and wrap the litellm.aspeech call with the max_parallel_requests semaphore plus async_routing_strategy_pre_call_checks so TTS respects rpm limits the same way the other router endpoints do Also add a unit test that exercises _aspeech directly and asserts the deployment metadata reaches the underlying call * fix(slack_alerting): stop false-positive hanging request alerts for requests below the alerting threshold (#30106) * fix(slack_alerting): skip hanging request alerts below the threshold The hanging request check alerted on any cached request whose completion status was not yet recorded, with no minimum age check. Since the background loop runs every alerting_threshold / 2 seconds, any request that happened to be in flight at a check fired a "hanging - Ns+ request time" alert even if it was only seconds old, producing a steady stream of false positives. Add a created_at timestamp to HangingRequestData, stamped when the request enters the hanging request cache, and skip requests younger than alerting_threshold without evicting them, so a later check can still alert if they never complete. Extend the cache TTL from threshold + 60s to 1.5x threshold + 60s; with the age check, entries only become alertable after threshold seconds, and the check period is threshold / 2, so the old TTL could evict a genuinely hanging request before any check saw it cross the threshold. Fixes #27855. * fix(slack_alerting): alert once per hanging request The min-age gate stops false positives for young in-flight requests, but a genuinely hanging request still re-alerted on every checker tick within the cache TTL. With the wider TTL (1.5x threshold + 60s) that is 1-2 extra Slack notifications per stuck request at the default 600s threshold. Flag a HangingRequestData entry as alerted once its alert fires and skip flagged entries on later ticks, so each hang produces exactly one alert. The cache reference is mutated in place, so the TTL is untouched and still handles cleanup. Adds a regression test asserting one alert across multiple ticks. Fixes #27855. * fix(health): treat all-proxy-models keys as unrestricted in /health (#30087) * fix(health): treat all-proxy-models keys as unrestricted in /health A key granted all model permissions stores the literal "all-proxy-models" marker in its models list. The /health access filter compared that marker against real model_names, so the model list filtered down to nothing and the WebUI health check returned healthy_count=0, unhealthy_count=0 with HTTP 503. Skip the filter (both the live path and the background-cache model_id scoping) when the marker is present, matching how auth_checks treats SpecialModelNames.all_proxy_models. Fixes #29744. * fix(health): resolve all-team-models sentinel to the team allowlist Same failure shape as the all-proxy-models case: a key carrying the literal "all-team-models" entry matches no real model_name, so the /health access filter would zero out the model list. Resolve the sentinel to the key's team models when team_id is set, matching get_key_models in model_checks.py. Without a team_id the sentinel stays unresolved and matches nothing, denying rather than widening access, mirroring _resolve_key_models_for_auth_check. * feat(proxy): auto-enable drop_params for Claude Code requests (#30218) * feat(proxy): auto-enable drop_params for Claude Code requests Claude Code identifies itself with a claude-cli/<version> user agent and sends Anthropic-specific params (top_k, thinking, etc.) on every request. When the proxy routes those requests to a non-Anthropic provider, the unsupported params fail the call unless drop_params is configured. Detect the Claude Code user agent in add_litellm_data_to_request and default drop_params to true for those requests, without overriding an explicit drop_params value sent by the caller. * feat(proxy): respect operator litellm_settings drop_params over Claude Code default An explicit drop_params in the operator's litellm_settings (true or false) now suppresses the Claude Code user agent default, so an operator who deliberately configured drop_params: false keeps strict param validation for Claude Code clients too. The auto-default only fills the gap when neither the request body nor the config sets a value. * fix(snowflake): migrate to native endpoints with auto-routing for Claude models (#29964) * fix(snowflake): migrate to native Cortex REST API endpoints Replaces the legacy /api/v2/cortex/inference:complete endpoint with the native OpenAI-compatible /api/v2/cortex/v1/chat/completions endpoint, fixing error 390142 (Incoming request does not contain a valid payload) when using model: snowflake/<model> in LiteLLM proxy. Changes: - litellm/llms/snowflake/chat/transformation.py: route to native /cortex/v1/chat/completions, remove Snowflake-specific tool_spec payload transformation, remove content_list response handling, add stream to supported params - litellm/llms/snowflake/anthropic/transformation.py (new): SnowflakeCortexAnthropicConfig routes Claude models to /cortex/v1/messages with anthropic-version header and Anthropic->OpenAI response transform - tests: 29 unit tests covering URL routing, auth headers, payload format, and response parsing * fix(snowflake): map max_tokens to max_completion_tokens for native endpoint * fix: handle multi-turn tool conversations and OpenAI→Anthropic tool format conversion - _extract_system_and_messages now preserves tool_calls from assistant messages and converts them to Anthropic tool_use content blocks - tool role messages are converted to user role with tool_result content blocks (as required by Anthropic Messages API) - Added _transform_tools_to_anthropic() to convert OpenAI tool format (type/function/parameters) to Anthropic format (name/input_schema) - Added comprehensive tests for multi-turn tool conversations Addresses review feedback on PR #29964 * test: add coverage for malformed JSON and non-string tool arguments * fix(tests): update chat transformation tests for native OpenAI-compatible endpoint * style: apply black formatting * fix: resolve mypy type errors in anthropic transformation * fix: correct mypy type: ignore error codes (attr-defined) * fix: use max_tokens instead of max_completion_tokens for Snowflake endpoint compatibility * refactor: merge Anthropic config into unified SnowflakeConfig with auto-routing - Remove separate SnowflakeCortexAnthropicConfig and anthropic/ directory - SnowflakeConfig now auto-routes based on model name: - Claude models → /messages endpoint (Anthropic format) - All others → /chat/completions endpoint (OpenAI format) - No new provider needed (stays as SNOWFLAKE = 'snowflake') - Tool message transformation for Claude: tool_calls → tool_use blocks, tool role → user with tool_result - OpenAI → Anthropic tool format conversion (parameters → input_schema) - Addresses Greptile feedback about unwired SnowflakeCortexAnthropicConfig * fix: use max_completion_tokens for /chat/completions (Snowflake deprecated max_tokens on this endpoint) * fix(tests): update assertions for Claude auto-routing to /messages endpoint * fix(snowflake): add tool_choice conversion and preserve max_completion_tokens in Anthropic path * fix(snowflake): use ChatCompletionMessageToolCall objects and strip model prefix on OpenAI path * fix(snowflake): collect multiple system messages to prevent guardrail override * chore: remove committed .pyc files and add __pycache__ to .gitignore * fix: remove unused Union import * fix: restore original .gitignore (accidentally replaced in earlier commit) * feat(snowflake): add streaming response handler for both Anthropic and OpenAI SSE formats * fix: remove unused AsyncIterator and Iterator imports * fix: add missing total_tokens to ChatCompletionUsageBlock * fix(snowflake): coalesce consecutive tool results into single user message for Anthropic * fix(snowflake): handle message_start event for streaming input_tokens tracking * fix: evict last deleted model in multi-instance deployments (#28608) * fix: evict last deleted model in multi-instance deployments _delete_deployment had an early return when db_models was empty, preventing eviction of the last deleted model during reconciliation. - Remove len(db_models)==0 early return from _delete_deployment - Return None (not []) from _get_models_from_db on DB failure so callers can distinguish a transient failure from a genuinely empty DB - Guard _update_llm_router against None to skip updates on DB failure Fixes #28443 * test: remove dead MagicMock assignment in type_mismatch test * fix: update test to pass [] not None to _update_llm_router test_ProxyConfig__update_llm_router_bad_proxy_logging_raises was passing None as new_models to get through to the proxy_logging_obj check, but the None guard we added now returns early before reaching that path. Pass [] instead so the test exercises the intended AttributeError case. Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> * chore: regenerate API types to sync schema.d.ts with proxy OpenAPI spec Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> --------- Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> * fix: invalidate Redis spend counter on /key/reset_spend (#29694) * fix: set Redis spend counter to reset_to value on /key/reset_spend Previously, the Redis spend counter was always set to 0.0 after a reset, even when reset_to was a non-zero value (partial reset). This caused the budget to be under-enforced for up to 60 seconds until the counter expired and fell through to the DB. Now the counter is set to the actual reset_to value, so partial resets are reflected correctly and budget enforcement is consistent. * test: update reset_key_spend test to match direct cache set The implementation now sets spend_counter_cache directly instead of calling _invalidate_spend_counter. Update the test to verify the in_memory_cache.set_cache call with the correct key, value, and ttl. --------- Co-authored-by: michaelxer <michaelxer@users.noreply.github.com> * fix: add scaleway models pricing (#27659) * fix: Add embeddings support for Scaleway provider * fix: resolve merge conflicts * fix(main): clarify backend route handling for Swagger static assets (#30196) * fix(main): clarify backend route handling for Swagger static assets * fix(allowlist): add BACKEND_MOUNT_PATHS for Swagger static assets * fix(voyage): route multimodal embeddings to correct endpoint (#30193) * fix(voyage): route multimodal embeddings to correct endpoint * test(voyage): cover multimodal embedding edge cases * test(voyage): cover api key fallback * fix(voyage): raise early on missing api key and malformed image url * test(voyage): cover utils routing and helper * fix(voyage): route supported openai params for multimodal models * style: apply black formatting * fix(ui): infer Azure API version from API base (#30204) * fix(ui): infer Azure API version from API base * fix(ui): address Azure API version feedback * Update litellm/llms/snowflake/chat/transformation.py Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> * feat(datadog): add team-scoped Datadog callback support (#29947) Enable teams to configure their own Datadog credentials via POST /team/{team_id}/callback, following the same pattern as Langfuse. * Merge pull request #29528 from aanchal22/litellm_byok-alias-merge fix(proxy): atomic merge for team model aliases and team.models on BYOK create * feat: add EmpirioLabs as an OpenAI-compatible provider (#30278) Co-authored-by: Adam Dalloul <adam.d.developer@gmail.com> * fix: resolve failing tests and lint in snowflake/team endpoints - Black-format snowflake/chat/transformation.py to fix lint failure - Update Anthropic config test to expect default max_tokens of 4096 (matches implementation) - Add AsyncMock + execute_raw mock to team_model_add cache-refresh pin test - Add model_dump mock and patch cache/logging in test_uses_atomic_array_append_with_dedup Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * fix(test): update test_db_error_new_model_check for new _delete_deployment logic _delete_deployment no longer short-circuits on empty db_models — it now treats [] as a valid empty-DB state and proceeds to check config models. Mock get_config to return the two router deployments so they appear in combined_id_list and are protected, which matches the real-world scenario where a DB error occurs but the models are config-backed. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * feat(proxy): register cancel_on_disconnect in ConfigGeneralSettings and config list (#30295) * feat(proxy): register cancel_on_disconnect in ConfigGeneralSettings and config list Follow-up to #30223 per maintainer review: documents the flag in ConfigGeneralSettings with a short description and adds it to allowed_args in get_config_list so the UI and /config/list expose it. A test pins that /config/list returns the field with type Boolean, which requires both registrations to be present * chore(ui): regenerate schema.d.ts for cancel_on_disconnect --------- Co-authored-by: kursad <kursad.lacin@brado.net> * fix(datadog): never fall back to env DD_API_KEY for caller-supplied destinations Team/key-scoped Datadog loggers could be pointed at an arbitrary dd_agent_host or dd_site while omitting dd_api_key, causing the proxy's global DD_API_KEY to be sent as the DD-API-KEY header to that destination. Gate the env-var fallback behind an allow_env_credentials flag, set to False when the destination is caller-supplied, mirroring the existing langfuse/langsmith pattern. --------- Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> Co-authored-by: Emerson Gomes <emerson.gomes@thalesgroup.com> Co-authored-by: daitran-tensormesh <dai@tensormesh.ai> Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> Co-authored-by: Muspi Merol <me@promplate.dev> Co-authored-by: fangkang <fangkangm@gmail.com> Co-authored-by: Chris Hoogeboom <chris.hoogeboom@gmail.com> Co-authored-by: kursadlacin <kursadlacin@gmail.com> Co-authored-by: kursad <kursad.lacin@brado.net> Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> Co-authored-by: hcl <chenglunhu@gmail.com> Co-authored-by: Filippo Menghi <113345637+Cyberfilo@users.noreply.github.com> Co-authored-by: Mateo Wang <277851410+mateo-berri@users.noreply.github.com> Co-authored-by: sfc-gh-nashukla <navnit.shukla@snowflake.com> Co-authored-by: Rudra Dudhat <contact.rdudhat@gmail.com> Co-authored-by: Michael <52305679+michaelxer@users.noreply.github.com> Co-authored-by: michaelxer <michaelxer@users.noreply.github.com> Co-authored-by: Quentin Champenois <26109239+Quentinchampenois@users.noreply.github.com> Co-authored-by: mauriceberentsen <mauriceberentsen@live.nl> Co-authored-by: lost9999 <56498264+lost9999@users.noreply.github.com> Co-authored-by: GaetanVDB07 <86427581+GaetanVDB07@users.noreply.github.com> Co-authored-by: Aanchal Khandelwal <aan2210khandelwal@gmail.com> Co-authored-by: Adam Dalloul <adam_dalloul@icloud.com> Co-authored-by: Adam Dalloul <adam.d.developer@gmail.com> Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
michaelxer
added a commit
to michaelxer/litellm
that referenced
this pull request
Jun 17, 2026
* feat(bedrock): add bedrock mantle gemma 4 models (BerriAI#30264) * feat(bedrock): add bedrock mantle gemma 4 models * test(bedrock): harden mantle local cost fixture * feat(responses): enable the responses API for the Tensormesh provider (BerriAI#30209) * feat(responses): enable the responses API for the Tensormesh provider * Update litellm/llms/openai_like/providers.json Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> --------- Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> * fix(langfuse_otel): mark LLM spans as generations (BerriAI#30250) * fix(bedrock): stop stream_chunk_size leaking into invoke request bodies (BerriAI#30240) stream_chunk_size is a LiteLLM-internal knob for re-chunking the HTTP response stream. The invoke transformations splat optional_params into the provider request body without dropping it, and Bedrock rejects unknown fields, so any bedrock/invoke request that sets the parameter fails with ValidationException: stream_chunk_size: Extra inputs are not permitted. Drop it in the invoke dispatcher (covers cohere, titan, mistral, meta, ai21) and in the Claude messages-format request builder (the route used for bedrock/invoke Anthropic models) * fix(bedrock): stop buffering streamed tool-call argument deltas (BerriAI#30231) * fix(bedrock): stop buffering streamed tool-call argument deltas Two issues made Bedrock tool-use streaming arrive as a single end-of-stream burst through LiteLLM while plain text streamed fine. First, the anthropic-beta allowlist mapped fine-grained-tool-streaming-2025-05-14 to null for bedrock and bedrock_converse, so the header was silently stripped. Without that beta, Anthropic models on Bedrock buffer tool input server-side and emit all toolUse.input deltas at once (verified against converse-stream and invoke-with-response-stream directly). Bedrock accepts the beta via additionalModelRequestFields.anthropic_beta, so it is now forwarded. Second, the streaming reads re-chunked the AWS event stream with iter_bytes(chunk_size=1024). httpx's ByteChunker only releases full 1024-byte blocks, so the small early events (messageStart, contentBlockStart, first deltas) sat in the buffer until enough bytes accumulated, pushing time-to-first-byte from ~1.4s to ~8.5s on buffered tool-use streams. The default is now no re-chunking; an explicit stream_chunk_size is still honored. * test(bedrock): cover explicit stream_chunk_size on sync invoke path * test(bedrock): cover stream_chunk_size plumbing through converse completion * test(bedrock): cover stream_chunk_size default in legacy BedrockLLM streaming * test(bedrock): merge converse handler tests into existing mapped test file pytest imports test modules by basename in non-package test dirs, so the new tests/test_litellm/llms/bedrock/chat/test_converse_handler.py collided with the pre-existing tests/test_litellm/llms/chat/test_converse_handler.py and broke collection in CI. Move the new tests into the existing file * feat(otel): emit v2 cost breakdown + stamp tracer scope version (BerriAI#30156) Read the StandardLoggingPayload cost_breakdown into a typed LLMCost on LLMCallSpanData and emit each component under litellm.cost.* (absent components omitted, so spans stay sparse). Stamp litellm.__version__ as the instrumentation scope version so every v2 span carries a deterministic scope.version. Tests under tests/test_litellm/integrations/otel/. * fix(proxy): cancel in-flight upstream LLM request on client disconnect (opt-in) (BerriAI#30223) * fix(proxy): cancel in-flight upstream LLM request on client disconnect (opt-in) On the non-streaming path, base_process_llm_request awaited the LLM call with no disconnect monitoring; when the HTTP client went away the upstream request kept running until completion or request_timeout (6000s default), holding a backend slot (e.g. a vLLM GPU slot) for output nobody would read Add an opt-in general_settings.cancel_on_disconnect flag, default off, so the default code path is unchanged. When enabled, a receive-based watcher task observes http.disconnect and cancels the asyncio.gather driving the upstream call. The resulting CancelledError is converted to HTTPException 499 only when the disconnect event is set, so server-initiated cancellations still propagate as-is. The 499 then flows through _handle_llm_api_exception like any other failure, meaning post_call_failure_hook still releases max_parallel_requests slots and fires spend and alerting callbacks; it is logged at info level instead of a full traceback Also removes the dead check_request_disconnection helper in proxy_server.py (zero call sites) along with its behavior-pin tests Builds on the receive-based design from BerriAI#25776 Addresses BerriAI#13774. Re-fixes BerriAI#22805 (regressed after the BerriAI#14295 revert) Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> * fix(proxy): scope 499 quiet logging to disconnects and harden watcher Address the two P2 findings from the Greptile review on BerriAI#30223. The info-level logging in _log_llm_api_exception now applies only to the disconnect-specific HTTPException (status 499 plus the shared _CLIENT_DISCONNECT_DETAIL message), so any other 499 raised by hooks or guardrails keeps its full traceback. The disconnect watcher now catches exceptions from request.receive() (e.g. a transport reset) and logs a warning instead of dying silently, making the degradation to no-op visible; a test pins that the LLM call is not cancelled in that case --------- Co-authored-by: kursad <kursad.lacin@brado.net> Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> * fix(bedrock): grant aws-external-anthropic:* in OIDC session policy for claude_platform (BerriAI#30200) (BerriAI#30205) The inline STS session policy passed to assume_role_with_web_identity acts as an IAM PERMISSION CEILING — effective permissions are the intersection of the role's identity policies and this policy. Any action not listed is silently denied even when the IAM role grants it. BerriAI#27678 added the bedrock/claude_platform/<model> route but its service-side action namespace is aws-external-anthropic:*, not bedrock:*. Without a matching statement here, every claude_platform request via OIDC (GCP federation, EKS Pod Identity webhook, etc.) 403s with 'no session policy allows the aws-external-anthropic:CreateInference action' — even with a fully permissive identity policy. Add a second ClaudePlatformLiteLLM statement covering CreateInference, CreateBatchInference, CancelBatchInference, DeleteBatchInference, CountTokens, Get*, List*. Keep aws:SecureTransport=true parity with the bedrock statement. Static creds + IRSA flow through different code paths and are not affected. Fixes BerriAI#30200 * fix(proxy): set Retry-After header on RouterRateLimitError 429 responses (BerriAI#30098) * Set Retry-After header on RouterRateLimitError responses When all deployments for a model are in cooldown, the proxy returns a 429 whose cooldown timing is only available by parsing the error message string. RouterRateLimitError already carries cooldown_time, so expose it as a standard retry-after header in _handle_llm_api_exception. The value is rounded up so clients never retry before the cooldown window ends. Fixes BerriAI#27823. * Set Retry-After after response-headers hook so cooldown wins The cooldown-derived retry-after was assigned before the post_call_response_headers_hook merge, so a callback returning a retry-after key (including a stale or empty value) silently clobbered it. Move the RouterRateLimitError block after the callback merge so the cooldown value is authoritative for this error type. * fix(router): route aspeech through async_function_with_fallbacks (BerriAI#30104) * fix(router): route aspeech through async_function_with_fallbacks Router.aspeech selected a deployment and awaited litellm.aspeech directly, so TTS requests got no retry on failure and no failover to backup deployments; the except block only fired an exception alert and re-raised. Every other router endpoint (acompletion, aembedding, atranscription, arerank) already delegates to async_function_with_fallbacks Mirror the atranscription pattern: move deployment selection and the litellm.aspeech call into a private _aspeech method, then have the public aspeech set kwargs["original_function"] = self._aspeech and await self.async_function_with_fallbacks(**kwargs). _aspeech also picks up the shared _get_async_openai_model_client helper and the same total/success/fail call accounting the sibling endpoints use Fixes BerriAI#27778. * fix(router): apply deployment kwargs and rpm semaphore in _aspeech Bring _aspeech fully in line with _atranscription: call _update_kwargs_with_deployment so deployment metadata, model_info, timeout, and default litellm params flow into the request, and wrap the litellm.aspeech call with the max_parallel_requests semaphore plus async_routing_strategy_pre_call_checks so TTS respects rpm limits the same way the other router endpoints do Also add a unit test that exercises _aspeech directly and asserts the deployment metadata reaches the underlying call * fix(slack_alerting): stop false-positive hanging request alerts for requests below the alerting threshold (BerriAI#30106) * fix(slack_alerting): skip hanging request alerts below the threshold The hanging request check alerted on any cached request whose completion status was not yet recorded, with no minimum age check. Since the background loop runs every alerting_threshold / 2 seconds, any request that happened to be in flight at a check fired a "hanging - Ns+ request time" alert even if it was only seconds old, producing a steady stream of false positives. Add a created_at timestamp to HangingRequestData, stamped when the request enters the hanging request cache, and skip requests younger than alerting_threshold without evicting them, so a later check can still alert if they never complete. Extend the cache TTL from threshold + 60s to 1.5x threshold + 60s; with the age check, entries only become alertable after threshold seconds, and the check period is threshold / 2, so the old TTL could evict a genuinely hanging request before any check saw it cross the threshold. Fixes BerriAI#27855. * fix(slack_alerting): alert once per hanging request The min-age gate stops false positives for young in-flight requests, but a genuinely hanging request still re-alerted on every checker tick within the cache TTL. With the wider TTL (1.5x threshold + 60s) that is 1-2 extra Slack notifications per stuck request at the default 600s threshold. Flag a HangingRequestData entry as alerted once its alert fires and skip flagged entries on later ticks, so each hang produces exactly one alert. The cache reference is mutated in place, so the TTL is untouched and still handles cleanup. Adds a regression test asserting one alert across multiple ticks. Fixes BerriAI#27855. * fix(health): treat all-proxy-models keys as unrestricted in /health (BerriAI#30087) * fix(health): treat all-proxy-models keys as unrestricted in /health A key granted all model permissions stores the literal "all-proxy-models" marker in its models list. The /health access filter compared that marker against real model_names, so the model list filtered down to nothing and the WebUI health check returned healthy_count=0, unhealthy_count=0 with HTTP 503. Skip the filter (both the live path and the background-cache model_id scoping) when the marker is present, matching how auth_checks treats SpecialModelNames.all_proxy_models. Fixes BerriAI#29744. * fix(health): resolve all-team-models sentinel to the team allowlist Same failure shape as the all-proxy-models case: a key carrying the literal "all-team-models" entry matches no real model_name, so the /health access filter would zero out the model list. Resolve the sentinel to the key's team models when team_id is set, matching get_key_models in model_checks.py. Without a team_id the sentinel stays unresolved and matches nothing, denying rather than widening access, mirroring _resolve_key_models_for_auth_check. * feat(proxy): auto-enable drop_params for Claude Code requests (BerriAI#30218) * feat(proxy): auto-enable drop_params for Claude Code requests Claude Code identifies itself with a claude-cli/<version> user agent and sends Anthropic-specific params (top_k, thinking, etc.) on every request. When the proxy routes those requests to a non-Anthropic provider, the unsupported params fail the call unless drop_params is configured. Detect the Claude Code user agent in add_litellm_data_to_request and default drop_params to true for those requests, without overriding an explicit drop_params value sent by the caller. * feat(proxy): respect operator litellm_settings drop_params over Claude Code default An explicit drop_params in the operator's litellm_settings (true or false) now suppresses the Claude Code user agent default, so an operator who deliberately configured drop_params: false keeps strict param validation for Claude Code clients too. The auto-default only fills the gap when neither the request body nor the config sets a value. * fix(snowflake): migrate to native endpoints with auto-routing for Claude models (BerriAI#29964) * fix(snowflake): migrate to native Cortex REST API endpoints Replaces the legacy /api/v2/cortex/inference:complete endpoint with the native OpenAI-compatible /api/v2/cortex/v1/chat/completions endpoint, fixing error 390142 (Incoming request does not contain a valid payload) when using model: snowflake/<model> in LiteLLM proxy. Changes: - litellm/llms/snowflake/chat/transformation.py: route to native /cortex/v1/chat/completions, remove Snowflake-specific tool_spec payload transformation, remove content_list response handling, add stream to supported params - litellm/llms/snowflake/anthropic/transformation.py (new): SnowflakeCortexAnthropicConfig routes Claude models to /cortex/v1/messages with anthropic-version header and Anthropic->OpenAI response transform - tests: 29 unit tests covering URL routing, auth headers, payload format, and response parsing * fix(snowflake): map max_tokens to max_completion_tokens for native endpoint * fix: handle multi-turn tool conversations and OpenAI→Anthropic tool format conversion - _extract_system_and_messages now preserves tool_calls from assistant messages and converts them to Anthropic tool_use content blocks - tool role messages are converted to user role with tool_result content blocks (as required by Anthropic Messages API) - Added _transform_tools_to_anthropic() to convert OpenAI tool format (type/function/parameters) to Anthropic format (name/input_schema) - Added comprehensive tests for multi-turn tool conversations Addresses review feedback on PR BerriAI#29964 * test: add coverage for malformed JSON and non-string tool arguments * fix(tests): update chat transformation tests for native OpenAI-compatible endpoint * style: apply black formatting * fix: resolve mypy type errors in anthropic transformation * fix: correct mypy type: ignore error codes (attr-defined) * fix: use max_tokens instead of max_completion_tokens for Snowflake endpoint compatibility * refactor: merge Anthropic config into unified SnowflakeConfig with auto-routing - Remove separate SnowflakeCortexAnthropicConfig and anthropic/ directory - SnowflakeConfig now auto-routes based on model name: - Claude models → /messages endpoint (Anthropic format) - All others → /chat/completions endpoint (OpenAI format) - No new provider needed (stays as SNOWFLAKE = 'snowflake') - Tool message transformation for Claude: tool_calls → tool_use blocks, tool role → user with tool_result - OpenAI → Anthropic tool format conversion (parameters → input_schema) - Addresses Greptile feedback about unwired SnowflakeCortexAnthropicConfig * fix: use max_completion_tokens for /chat/completions (Snowflake deprecated max_tokens on this endpoint) * fix(tests): update assertions for Claude auto-routing to /messages endpoint * fix(snowflake): add tool_choice conversion and preserve max_completion_tokens in Anthropic path * fix(snowflake): use ChatCompletionMessageToolCall objects and strip model prefix on OpenAI path * fix(snowflake): collect multiple system messages to prevent guardrail override * chore: remove committed .pyc files and add __pycache__ to .gitignore * fix: remove unused Union import * fix: restore original .gitignore (accidentally replaced in earlier commit) * feat(snowflake): add streaming response handler for both Anthropic and OpenAI SSE formats * fix: remove unused AsyncIterator and Iterator imports * fix: add missing total_tokens to ChatCompletionUsageBlock * fix(snowflake): coalesce consecutive tool results into single user message for Anthropic * fix(snowflake): handle message_start event for streaming input_tokens tracking * fix: evict last deleted model in multi-instance deployments (BerriAI#28608) * fix: evict last deleted model in multi-instance deployments _delete_deployment had an early return when db_models was empty, preventing eviction of the last deleted model during reconciliation. - Remove len(db_models)==0 early return from _delete_deployment - Return None (not []) from _get_models_from_db on DB failure so callers can distinguish a transient failure from a genuinely empty DB - Guard _update_llm_router against None to skip updates on DB failure Fixes BerriAI#28443 * test: remove dead MagicMock assignment in type_mismatch test * fix: update test to pass [] not None to _update_llm_router test_ProxyConfig__update_llm_router_bad_proxy_logging_raises was passing None as new_models to get through to the proxy_logging_obj check, but the None guard we added now returns early before reaching that path. Pass [] instead so the test exercises the intended AttributeError case. Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> * chore: regenerate API types to sync schema.d.ts with proxy OpenAPI spec Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> --------- Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> * fix: invalidate Redis spend counter on /key/reset_spend (BerriAI#29694) * fix: set Redis spend counter to reset_to value on /key/reset_spend Previously, the Redis spend counter was always set to 0.0 after a reset, even when reset_to was a non-zero value (partial reset). This caused the budget to be under-enforced for up to 60 seconds until the counter expired and fell through to the DB. Now the counter is set to the actual reset_to value, so partial resets are reflected correctly and budget enforcement is consistent. * test: update reset_key_spend test to match direct cache set The implementation now sets spend_counter_cache directly instead of calling _invalidate_spend_counter. Update the test to verify the in_memory_cache.set_cache call with the correct key, value, and ttl. --------- Co-authored-by: michaelxer <michaelxer@users.noreply.github.com> * fix: add scaleway models pricing (BerriAI#27659) * fix: Add embeddings support for Scaleway provider * fix: resolve merge conflicts * fix(main): clarify backend route handling for Swagger static assets (BerriAI#30196) * fix(main): clarify backend route handling for Swagger static assets * fix(allowlist): add BACKEND_MOUNT_PATHS for Swagger static assets * fix(voyage): route multimodal embeddings to correct endpoint (BerriAI#30193) * fix(voyage): route multimodal embeddings to correct endpoint * test(voyage): cover multimodal embedding edge cases * test(voyage): cover api key fallback * fix(voyage): raise early on missing api key and malformed image url * test(voyage): cover utils routing and helper * fix(voyage): route supported openai params for multimodal models * style: apply black formatting * fix(ui): infer Azure API version from API base (BerriAI#30204) * fix(ui): infer Azure API version from API base * fix(ui): address Azure API version feedback * Update litellm/llms/snowflake/chat/transformation.py Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> * feat(datadog): add team-scoped Datadog callback support (BerriAI#29947) Enable teams to configure their own Datadog credentials via POST /team/{team_id}/callback, following the same pattern as Langfuse. * Merge pull request BerriAI#29528 from aanchal22/litellm_byok-alias-merge fix(proxy): atomic merge for team model aliases and team.models on BYOK create * feat: add EmpirioLabs as an OpenAI-compatible provider (BerriAI#30278) Co-authored-by: Adam Dalloul <adam.d.developer@gmail.com> * fix: resolve failing tests and lint in snowflake/team endpoints - Black-format snowflake/chat/transformation.py to fix lint failure - Update Anthropic config test to expect default max_tokens of 4096 (matches implementation) - Add AsyncMock + execute_raw mock to team_model_add cache-refresh pin test - Add model_dump mock and patch cache/logging in test_uses_atomic_array_append_with_dedup Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * fix(test): update test_db_error_new_model_check for new _delete_deployment logic _delete_deployment no longer short-circuits on empty db_models — it now treats [] as a valid empty-DB state and proceeds to check config models. Mock get_config to return the two router deployments so they appear in combined_id_list and are protected, which matches the real-world scenario where a DB error occurs but the models are config-backed. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * feat(proxy): register cancel_on_disconnect in ConfigGeneralSettings and config list (BerriAI#30295) * feat(proxy): register cancel_on_disconnect in ConfigGeneralSettings and config list Follow-up to BerriAI#30223 per maintainer review: documents the flag in ConfigGeneralSettings with a short description and adds it to allowed_args in get_config_list so the UI and /config/list expose it. A test pins that /config/list returns the field with type Boolean, which requires both registrations to be present * chore(ui): regenerate schema.d.ts for cancel_on_disconnect --------- Co-authored-by: kursad <kursad.lacin@brado.net> * fix(datadog): never fall back to env DD_API_KEY for caller-supplied destinations Team/key-scoped Datadog loggers could be pointed at an arbitrary dd_agent_host or dd_site while omitting dd_api_key, causing the proxy's global DD_API_KEY to be sent as the DD-API-KEY header to that destination. Gate the env-var fallback behind an allow_env_credentials flag, set to False when the destination is caller-supplied, mirroring the existing langfuse/langsmith pattern. --------- Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> Co-authored-by: Emerson Gomes <emerson.gomes@thalesgroup.com> Co-authored-by: daitran-tensormesh <dai@tensormesh.ai> Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> Co-authored-by: Muspi Merol <me@promplate.dev> Co-authored-by: fangkang <fangkangm@gmail.com> Co-authored-by: Chris Hoogeboom <chris.hoogeboom@gmail.com> Co-authored-by: kursadlacin <kursadlacin@gmail.com> Co-authored-by: kursad <kursad.lacin@brado.net> Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> Co-authored-by: hcl <chenglunhu@gmail.com> Co-authored-by: Filippo Menghi <113345637+Cyberfilo@users.noreply.github.com> Co-authored-by: Mateo Wang <277851410+mateo-berri@users.noreply.github.com> Co-authored-by: sfc-gh-nashukla <navnit.shukla@snowflake.com> Co-authored-by: Rudra Dudhat <contact.rdudhat@gmail.com> Co-authored-by: Michael <52305679+michaelxer@users.noreply.github.com> Co-authored-by: michaelxer <michaelxer@users.noreply.github.com> Co-authored-by: Quentin Champenois <26109239+Quentinchampenois@users.noreply.github.com> Co-authored-by: mauriceberentsen <mauriceberentsen@live.nl> Co-authored-by: lost9999 <56498264+lost9999@users.noreply.github.com> Co-authored-by: GaetanVDB07 <86427581+GaetanVDB07@users.noreply.github.com> Co-authored-by: Aanchal Khandelwal <aan2210khandelwal@gmail.com> Co-authored-by: Adam Dalloul <adam_dalloul@icloud.com> Co-authored-by: Adam Dalloul <adam.d.developer@gmail.com> Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
michaelxer
added a commit
to michaelxer/litellm
that referenced
this pull request
Jun 17, 2026
* feat(bedrock): add bedrock mantle gemma 4 models (BerriAI#30264) * feat(bedrock): add bedrock mantle gemma 4 models * test(bedrock): harden mantle local cost fixture * feat(responses): enable the responses API for the Tensormesh provider (BerriAI#30209) * feat(responses): enable the responses API for the Tensormesh provider * Update litellm/llms/openai_like/providers.json Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> --------- Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> * fix(langfuse_otel): mark LLM spans as generations (BerriAI#30250) * fix(bedrock): stop stream_chunk_size leaking into invoke request bodies (BerriAI#30240) stream_chunk_size is a LiteLLM-internal knob for re-chunking the HTTP response stream. The invoke transformations splat optional_params into the provider request body without dropping it, and Bedrock rejects unknown fields, so any bedrock/invoke request that sets the parameter fails with ValidationException: stream_chunk_size: Extra inputs are not permitted. Drop it in the invoke dispatcher (covers cohere, titan, mistral, meta, ai21) and in the Claude messages-format request builder (the route used for bedrock/invoke Anthropic models) * fix(bedrock): stop buffering streamed tool-call argument deltas (BerriAI#30231) * fix(bedrock): stop buffering streamed tool-call argument deltas Two issues made Bedrock tool-use streaming arrive as a single end-of-stream burst through LiteLLM while plain text streamed fine. First, the anthropic-beta allowlist mapped fine-grained-tool-streaming-2025-05-14 to null for bedrock and bedrock_converse, so the header was silently stripped. Without that beta, Anthropic models on Bedrock buffer tool input server-side and emit all toolUse.input deltas at once (verified against converse-stream and invoke-with-response-stream directly). Bedrock accepts the beta via additionalModelRequestFields.anthropic_beta, so it is now forwarded. Second, the streaming reads re-chunked the AWS event stream with iter_bytes(chunk_size=1024). httpx's ByteChunker only releases full 1024-byte blocks, so the small early events (messageStart, contentBlockStart, first deltas) sat in the buffer until enough bytes accumulated, pushing time-to-first-byte from ~1.4s to ~8.5s on buffered tool-use streams. The default is now no re-chunking; an explicit stream_chunk_size is still honored. * test(bedrock): cover explicit stream_chunk_size on sync invoke path * test(bedrock): cover stream_chunk_size plumbing through converse completion * test(bedrock): cover stream_chunk_size default in legacy BedrockLLM streaming * test(bedrock): merge converse handler tests into existing mapped test file pytest imports test modules by basename in non-package test dirs, so the new tests/test_litellm/llms/bedrock/chat/test_converse_handler.py collided with the pre-existing tests/test_litellm/llms/chat/test_converse_handler.py and broke collection in CI. Move the new tests into the existing file * feat(otel): emit v2 cost breakdown + stamp tracer scope version (BerriAI#30156) Read the StandardLoggingPayload cost_breakdown into a typed LLMCost on LLMCallSpanData and emit each component under litellm.cost.* (absent components omitted, so spans stay sparse). Stamp litellm.__version__ as the instrumentation scope version so every v2 span carries a deterministic scope.version. Tests under tests/test_litellm/integrations/otel/. * fix(proxy): cancel in-flight upstream LLM request on client disconnect (opt-in) (BerriAI#30223) * fix(proxy): cancel in-flight upstream LLM request on client disconnect (opt-in) On the non-streaming path, base_process_llm_request awaited the LLM call with no disconnect monitoring; when the HTTP client went away the upstream request kept running until completion or request_timeout (6000s default), holding a backend slot (e.g. a vLLM GPU slot) for output nobody would read Add an opt-in general_settings.cancel_on_disconnect flag, default off, so the default code path is unchanged. When enabled, a receive-based watcher task observes http.disconnect and cancels the asyncio.gather driving the upstream call. The resulting CancelledError is converted to HTTPException 499 only when the disconnect event is set, so server-initiated cancellations still propagate as-is. The 499 then flows through _handle_llm_api_exception like any other failure, meaning post_call_failure_hook still releases max_parallel_requests slots and fires spend and alerting callbacks; it is logged at info level instead of a full traceback Also removes the dead check_request_disconnection helper in proxy_server.py (zero call sites) along with its behavior-pin tests Builds on the receive-based design from BerriAI#25776 Addresses BerriAI#13774. Re-fixes BerriAI#22805 (regressed after the BerriAI#14295 revert) Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> * fix(proxy): scope 499 quiet logging to disconnects and harden watcher Address the two P2 findings from the Greptile review on BerriAI#30223. The info-level logging in _log_llm_api_exception now applies only to the disconnect-specific HTTPException (status 499 plus the shared _CLIENT_DISCONNECT_DETAIL message), so any other 499 raised by hooks or guardrails keeps its full traceback. The disconnect watcher now catches exceptions from request.receive() (e.g. a transport reset) and logs a warning instead of dying silently, making the degradation to no-op visible; a test pins that the LLM call is not cancelled in that case --------- Co-authored-by: kursad <kursad.lacin@brado.net> Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> * fix(bedrock): grant aws-external-anthropic:* in OIDC session policy for claude_platform (BerriAI#30200) (BerriAI#30205) The inline STS session policy passed to assume_role_with_web_identity acts as an IAM PERMISSION CEILING — effective permissions are the intersection of the role's identity policies and this policy. Any action not listed is silently denied even when the IAM role grants it. BerriAI#27678 added the bedrock/claude_platform/<model> route but its service-side action namespace is aws-external-anthropic:*, not bedrock:*. Without a matching statement here, every claude_platform request via OIDC (GCP federation, EKS Pod Identity webhook, etc.) 403s with 'no session policy allows the aws-external-anthropic:CreateInference action' — even with a fully permissive identity policy. Add a second ClaudePlatformLiteLLM statement covering CreateInference, CreateBatchInference, CancelBatchInference, DeleteBatchInference, CountTokens, Get*, List*. Keep aws:SecureTransport=true parity with the bedrock statement. Static creds + IRSA flow through different code paths and are not affected. Fixes BerriAI#30200 * fix(proxy): set Retry-After header on RouterRateLimitError 429 responses (BerriAI#30098) * Set Retry-After header on RouterRateLimitError responses When all deployments for a model are in cooldown, the proxy returns a 429 whose cooldown timing is only available by parsing the error message string. RouterRateLimitError already carries cooldown_time, so expose it as a standard retry-after header in _handle_llm_api_exception. The value is rounded up so clients never retry before the cooldown window ends. Fixes BerriAI#27823. * Set Retry-After after response-headers hook so cooldown wins The cooldown-derived retry-after was assigned before the post_call_response_headers_hook merge, so a callback returning a retry-after key (including a stale or empty value) silently clobbered it. Move the RouterRateLimitError block after the callback merge so the cooldown value is authoritative for this error type. * fix(router): route aspeech through async_function_with_fallbacks (BerriAI#30104) * fix(router): route aspeech through async_function_with_fallbacks Router.aspeech selected a deployment and awaited litellm.aspeech directly, so TTS requests got no retry on failure and no failover to backup deployments; the except block only fired an exception alert and re-raised. Every other router endpoint (acompletion, aembedding, atranscription, arerank) already delegates to async_function_with_fallbacks Mirror the atranscription pattern: move deployment selection and the litellm.aspeech call into a private _aspeech method, then have the public aspeech set kwargs["original_function"] = self._aspeech and await self.async_function_with_fallbacks(**kwargs). _aspeech also picks up the shared _get_async_openai_model_client helper and the same total/success/fail call accounting the sibling endpoints use Fixes BerriAI#27778. * fix(router): apply deployment kwargs and rpm semaphore in _aspeech Bring _aspeech fully in line with _atranscription: call _update_kwargs_with_deployment so deployment metadata, model_info, timeout, and default litellm params flow into the request, and wrap the litellm.aspeech call with the max_parallel_requests semaphore plus async_routing_strategy_pre_call_checks so TTS respects rpm limits the same way the other router endpoints do Also add a unit test that exercises _aspeech directly and asserts the deployment metadata reaches the underlying call * fix(slack_alerting): stop false-positive hanging request alerts for requests below the alerting threshold (BerriAI#30106) * fix(slack_alerting): skip hanging request alerts below the threshold The hanging request check alerted on any cached request whose completion status was not yet recorded, with no minimum age check. Since the background loop runs every alerting_threshold / 2 seconds, any request that happened to be in flight at a check fired a "hanging - Ns+ request time" alert even if it was only seconds old, producing a steady stream of false positives. Add a created_at timestamp to HangingRequestData, stamped when the request enters the hanging request cache, and skip requests younger than alerting_threshold without evicting them, so a later check can still alert if they never complete. Extend the cache TTL from threshold + 60s to 1.5x threshold + 60s; with the age check, entries only become alertable after threshold seconds, and the check period is threshold / 2, so the old TTL could evict a genuinely hanging request before any check saw it cross the threshold. Fixes BerriAI#27855. * fix(slack_alerting): alert once per hanging request The min-age gate stops false positives for young in-flight requests, but a genuinely hanging request still re-alerted on every checker tick within the cache TTL. With the wider TTL (1.5x threshold + 60s) that is 1-2 extra Slack notifications per stuck request at the default 600s threshold. Flag a HangingRequestData entry as alerted once its alert fires and skip flagged entries on later ticks, so each hang produces exactly one alert. The cache reference is mutated in place, so the TTL is untouched and still handles cleanup. Adds a regression test asserting one alert across multiple ticks. Fixes BerriAI#27855. * fix(health): treat all-proxy-models keys as unrestricted in /health (BerriAI#30087) * fix(health): treat all-proxy-models keys as unrestricted in /health A key granted all model permissions stores the literal "all-proxy-models" marker in its models list. The /health access filter compared that marker against real model_names, so the model list filtered down to nothing and the WebUI health check returned healthy_count=0, unhealthy_count=0 with HTTP 503. Skip the filter (both the live path and the background-cache model_id scoping) when the marker is present, matching how auth_checks treats SpecialModelNames.all_proxy_models. Fixes BerriAI#29744. * fix(health): resolve all-team-models sentinel to the team allowlist Same failure shape as the all-proxy-models case: a key carrying the literal "all-team-models" entry matches no real model_name, so the /health access filter would zero out the model list. Resolve the sentinel to the key's team models when team_id is set, matching get_key_models in model_checks.py. Without a team_id the sentinel stays unresolved and matches nothing, denying rather than widening access, mirroring _resolve_key_models_for_auth_check. * feat(proxy): auto-enable drop_params for Claude Code requests (BerriAI#30218) * feat(proxy): auto-enable drop_params for Claude Code requests Claude Code identifies itself with a claude-cli/<version> user agent and sends Anthropic-specific params (top_k, thinking, etc.) on every request. When the proxy routes those requests to a non-Anthropic provider, the unsupported params fail the call unless drop_params is configured. Detect the Claude Code user agent in add_litellm_data_to_request and default drop_params to true for those requests, without overriding an explicit drop_params value sent by the caller. * feat(proxy): respect operator litellm_settings drop_params over Claude Code default An explicit drop_params in the operator's litellm_settings (true or false) now suppresses the Claude Code user agent default, so an operator who deliberately configured drop_params: false keeps strict param validation for Claude Code clients too. The auto-default only fills the gap when neither the request body nor the config sets a value. * fix(snowflake): migrate to native endpoints with auto-routing for Claude models (BerriAI#29964) * fix(snowflake): migrate to native Cortex REST API endpoints Replaces the legacy /api/v2/cortex/inference:complete endpoint with the native OpenAI-compatible /api/v2/cortex/v1/chat/completions endpoint, fixing error 390142 (Incoming request does not contain a valid payload) when using model: snowflake/<model> in LiteLLM proxy. Changes: - litellm/llms/snowflake/chat/transformation.py: route to native /cortex/v1/chat/completions, remove Snowflake-specific tool_spec payload transformation, remove content_list response handling, add stream to supported params - litellm/llms/snowflake/anthropic/transformation.py (new): SnowflakeCortexAnthropicConfig routes Claude models to /cortex/v1/messages with anthropic-version header and Anthropic->OpenAI response transform - tests: 29 unit tests covering URL routing, auth headers, payload format, and response parsing * fix(snowflake): map max_tokens to max_completion_tokens for native endpoint * fix: handle multi-turn tool conversations and OpenAI→Anthropic tool format conversion - _extract_system_and_messages now preserves tool_calls from assistant messages and converts them to Anthropic tool_use content blocks - tool role messages are converted to user role with tool_result content blocks (as required by Anthropic Messages API) - Added _transform_tools_to_anthropic() to convert OpenAI tool format (type/function/parameters) to Anthropic format (name/input_schema) - Added comprehensive tests for multi-turn tool conversations Addresses review feedback on PR BerriAI#29964 * test: add coverage for malformed JSON and non-string tool arguments * fix(tests): update chat transformation tests for native OpenAI-compatible endpoint * style: apply black formatting * fix: resolve mypy type errors in anthropic transformation * fix: correct mypy type: ignore error codes (attr-defined) * fix: use max_tokens instead of max_completion_tokens for Snowflake endpoint compatibility * refactor: merge Anthropic config into unified SnowflakeConfig with auto-routing - Remove separate SnowflakeCortexAnthropicConfig and anthropic/ directory - SnowflakeConfig now auto-routes based on model name: - Claude models → /messages endpoint (Anthropic format) - All others → /chat/completions endpoint (OpenAI format) - No new provider needed (stays as SNOWFLAKE = 'snowflake') - Tool message transformation for Claude: tool_calls → tool_use blocks, tool role → user with tool_result - OpenAI → Anthropic tool format conversion (parameters → input_schema) - Addresses Greptile feedback about unwired SnowflakeCortexAnthropicConfig * fix: use max_completion_tokens for /chat/completions (Snowflake deprecated max_tokens on this endpoint) * fix(tests): update assertions for Claude auto-routing to /messages endpoint * fix(snowflake): add tool_choice conversion and preserve max_completion_tokens in Anthropic path * fix(snowflake): use ChatCompletionMessageToolCall objects and strip model prefix on OpenAI path * fix(snowflake): collect multiple system messages to prevent guardrail override * chore: remove committed .pyc files and add __pycache__ to .gitignore * fix: remove unused Union import * fix: restore original .gitignore (accidentally replaced in earlier commit) * feat(snowflake): add streaming response handler for both Anthropic and OpenAI SSE formats * fix: remove unused AsyncIterator and Iterator imports * fix: add missing total_tokens to ChatCompletionUsageBlock * fix(snowflake): coalesce consecutive tool results into single user message for Anthropic * fix(snowflake): handle message_start event for streaming input_tokens tracking * fix: evict last deleted model in multi-instance deployments (BerriAI#28608) * fix: evict last deleted model in multi-instance deployments _delete_deployment had an early return when db_models was empty, preventing eviction of the last deleted model during reconciliation. - Remove len(db_models)==0 early return from _delete_deployment - Return None (not []) from _get_models_from_db on DB failure so callers can distinguish a transient failure from a genuinely empty DB - Guard _update_llm_router against None to skip updates on DB failure Fixes BerriAI#28443 * test: remove dead MagicMock assignment in type_mismatch test * fix: update test to pass [] not None to _update_llm_router test_ProxyConfig__update_llm_router_bad_proxy_logging_raises was passing None as new_models to get through to the proxy_logging_obj check, but the None guard we added now returns early before reaching that path. Pass [] instead so the test exercises the intended AttributeError case. Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> * chore: regenerate API types to sync schema.d.ts with proxy OpenAPI spec Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> --------- Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> * fix: invalidate Redis spend counter on /key/reset_spend (BerriAI#29694) * fix: set Redis spend counter to reset_to value on /key/reset_spend Previously, the Redis spend counter was always set to 0.0 after a reset, even when reset_to was a non-zero value (partial reset). This caused the budget to be under-enforced for up to 60 seconds until the counter expired and fell through to the DB. Now the counter is set to the actual reset_to value, so partial resets are reflected correctly and budget enforcement is consistent. * test: update reset_key_spend test to match direct cache set The implementation now sets spend_counter_cache directly instead of calling _invalidate_spend_counter. Update the test to verify the in_memory_cache.set_cache call with the correct key, value, and ttl. --------- Co-authored-by: michaelxer <michaelxer@users.noreply.github.com> * fix: add scaleway models pricing (BerriAI#27659) * fix: Add embeddings support for Scaleway provider * fix: resolve merge conflicts * fix(main): clarify backend route handling for Swagger static assets (BerriAI#30196) * fix(main): clarify backend route handling for Swagger static assets * fix(allowlist): add BACKEND_MOUNT_PATHS for Swagger static assets * fix(voyage): route multimodal embeddings to correct endpoint (BerriAI#30193) * fix(voyage): route multimodal embeddings to correct endpoint * test(voyage): cover multimodal embedding edge cases * test(voyage): cover api key fallback * fix(voyage): raise early on missing api key and malformed image url * test(voyage): cover utils routing and helper * fix(voyage): route supported openai params for multimodal models * style: apply black formatting * fix(ui): infer Azure API version from API base (BerriAI#30204) * fix(ui): infer Azure API version from API base * fix(ui): address Azure API version feedback * Update litellm/llms/snowflake/chat/transformation.py Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> * feat(datadog): add team-scoped Datadog callback support (BerriAI#29947) Enable teams to configure their own Datadog credentials via POST /team/{team_id}/callback, following the same pattern as Langfuse. * Merge pull request BerriAI#29528 from aanchal22/litellm_byok-alias-merge fix(proxy): atomic merge for team model aliases and team.models on BYOK create * feat: add EmpirioLabs as an OpenAI-compatible provider (BerriAI#30278) Co-authored-by: Adam Dalloul <adam.d.developer@gmail.com> * fix: resolve failing tests and lint in snowflake/team endpoints - Black-format snowflake/chat/transformation.py to fix lint failure - Update Anthropic config test to expect default max_tokens of 4096 (matches implementation) - Add AsyncMock + execute_raw mock to team_model_add cache-refresh pin test - Add model_dump mock and patch cache/logging in test_uses_atomic_array_append_with_dedup Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * fix(test): update test_db_error_new_model_check for new _delete_deployment logic _delete_deployment no longer short-circuits on empty db_models — it now treats [] as a valid empty-DB state and proceeds to check config models. Mock get_config to return the two router deployments so they appear in combined_id_list and are protected, which matches the real-world scenario where a DB error occurs but the models are config-backed. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * feat(proxy): register cancel_on_disconnect in ConfigGeneralSettings and config list (BerriAI#30295) * feat(proxy): register cancel_on_disconnect in ConfigGeneralSettings and config list Follow-up to BerriAI#30223 per maintainer review: documents the flag in ConfigGeneralSettings with a short description and adds it to allowed_args in get_config_list so the UI and /config/list expose it. A test pins that /config/list returns the field with type Boolean, which requires both registrations to be present * chore(ui): regenerate schema.d.ts for cancel_on_disconnect --------- Co-authored-by: kursad <kursad.lacin@brado.net> * fix(datadog): never fall back to env DD_API_KEY for caller-supplied destinations Team/key-scoped Datadog loggers could be pointed at an arbitrary dd_agent_host or dd_site while omitting dd_api_key, causing the proxy's global DD_API_KEY to be sent as the DD-API-KEY header to that destination. Gate the env-var fallback behind an allow_env_credentials flag, set to False when the destination is caller-supplied, mirroring the existing langfuse/langsmith pattern. --------- Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> Co-authored-by: Emerson Gomes <emerson.gomes@thalesgroup.com> Co-authored-by: daitran-tensormesh <dai@tensormesh.ai> Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> Co-authored-by: Muspi Merol <me@promplate.dev> Co-authored-by: fangkang <fangkangm@gmail.com> Co-authored-by: Chris Hoogeboom <chris.hoogeboom@gmail.com> Co-authored-by: kursadlacin <kursadlacin@gmail.com> Co-authored-by: kursad <kursad.lacin@brado.net> Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> Co-authored-by: hcl <chenglunhu@gmail.com> Co-authored-by: Filippo Menghi <113345637+Cyberfilo@users.noreply.github.com> Co-authored-by: Mateo Wang <277851410+mateo-berri@users.noreply.github.com> Co-authored-by: sfc-gh-nashukla <navnit.shukla@snowflake.com> Co-authored-by: Rudra Dudhat <contact.rdudhat@gmail.com> Co-authored-by: Michael <52305679+michaelxer@users.noreply.github.com> Co-authored-by: michaelxer <michaelxer@users.noreply.github.com> Co-authored-by: Quentin Champenois <26109239+Quentinchampenois@users.noreply.github.com> Co-authored-by: mauriceberentsen <mauriceberentsen@live.nl> Co-authored-by: lost9999 <56498264+lost9999@users.noreply.github.com> Co-authored-by: GaetanVDB07 <86427581+GaetanVDB07@users.noreply.github.com> Co-authored-by: Aanchal Khandelwal <aan2210khandelwal@gmail.com> Co-authored-by: Adam Dalloul <adam_dalloul@icloud.com> Co-authored-by: Adam Dalloul <adam.d.developer@gmail.com> Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
koladefaj
pushed a commit
to koladefaj/litellm
that referenced
this pull request
Jun 17, 2026
* feat(bedrock): add bedrock mantle gemma 4 models (BerriAI#30264) * feat(bedrock): add bedrock mantle gemma 4 models * test(bedrock): harden mantle local cost fixture * feat(responses): enable the responses API for the Tensormesh provider (BerriAI#30209) * feat(responses): enable the responses API for the Tensormesh provider * Update litellm/llms/openai_like/providers.json Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> --------- Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> * fix(langfuse_otel): mark LLM spans as generations (BerriAI#30250) * fix(bedrock): stop stream_chunk_size leaking into invoke request bodies (BerriAI#30240) stream_chunk_size is a LiteLLM-internal knob for re-chunking the HTTP response stream. The invoke transformations splat optional_params into the provider request body without dropping it, and Bedrock rejects unknown fields, so any bedrock/invoke request that sets the parameter fails with ValidationException: stream_chunk_size: Extra inputs are not permitted. Drop it in the invoke dispatcher (covers cohere, titan, mistral, meta, ai21) and in the Claude messages-format request builder (the route used for bedrock/invoke Anthropic models) * fix(bedrock): stop buffering streamed tool-call argument deltas (BerriAI#30231) * fix(bedrock): stop buffering streamed tool-call argument deltas Two issues made Bedrock tool-use streaming arrive as a single end-of-stream burst through LiteLLM while plain text streamed fine. First, the anthropic-beta allowlist mapped fine-grained-tool-streaming-2025-05-14 to null for bedrock and bedrock_converse, so the header was silently stripped. Without that beta, Anthropic models on Bedrock buffer tool input server-side and emit all toolUse.input deltas at once (verified against converse-stream and invoke-with-response-stream directly). Bedrock accepts the beta via additionalModelRequestFields.anthropic_beta, so it is now forwarded. Second, the streaming reads re-chunked the AWS event stream with iter_bytes(chunk_size=1024). httpx's ByteChunker only releases full 1024-byte blocks, so the small early events (messageStart, contentBlockStart, first deltas) sat in the buffer until enough bytes accumulated, pushing time-to-first-byte from ~1.4s to ~8.5s on buffered tool-use streams. The default is now no re-chunking; an explicit stream_chunk_size is still honored. * test(bedrock): cover explicit stream_chunk_size on sync invoke path * test(bedrock): cover stream_chunk_size plumbing through converse completion * test(bedrock): cover stream_chunk_size default in legacy BedrockLLM streaming * test(bedrock): merge converse handler tests into existing mapped test file pytest imports test modules by basename in non-package test dirs, so the new tests/test_litellm/llms/bedrock/chat/test_converse_handler.py collided with the pre-existing tests/test_litellm/llms/chat/test_converse_handler.py and broke collection in CI. Move the new tests into the existing file * feat(otel): emit v2 cost breakdown + stamp tracer scope version (BerriAI#30156) Read the StandardLoggingPayload cost_breakdown into a typed LLMCost on LLMCallSpanData and emit each component under litellm.cost.* (absent components omitted, so spans stay sparse). Stamp litellm.__version__ as the instrumentation scope version so every v2 span carries a deterministic scope.version. Tests under tests/test_litellm/integrations/otel/. * fix(proxy): cancel in-flight upstream LLM request on client disconnect (opt-in) (BerriAI#30223) * fix(proxy): cancel in-flight upstream LLM request on client disconnect (opt-in) On the non-streaming path, base_process_llm_request awaited the LLM call with no disconnect monitoring; when the HTTP client went away the upstream request kept running until completion or request_timeout (6000s default), holding a backend slot (e.g. a vLLM GPU slot) for output nobody would read Add an opt-in general_settings.cancel_on_disconnect flag, default off, so the default code path is unchanged. When enabled, a receive-based watcher task observes http.disconnect and cancels the asyncio.gather driving the upstream call. The resulting CancelledError is converted to HTTPException 499 only when the disconnect event is set, so server-initiated cancellations still propagate as-is. The 499 then flows through _handle_llm_api_exception like any other failure, meaning post_call_failure_hook still releases max_parallel_requests slots and fires spend and alerting callbacks; it is logged at info level instead of a full traceback Also removes the dead check_request_disconnection helper in proxy_server.py (zero call sites) along with its behavior-pin tests Builds on the receive-based design from BerriAI#25776 Addresses BerriAI#13774. Re-fixes BerriAI#22805 (regressed after the BerriAI#14295 revert) Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> * fix(proxy): scope 499 quiet logging to disconnects and harden watcher Address the two P2 findings from the Greptile review on BerriAI#30223. The info-level logging in _log_llm_api_exception now applies only to the disconnect-specific HTTPException (status 499 plus the shared _CLIENT_DISCONNECT_DETAIL message), so any other 499 raised by hooks or guardrails keeps its full traceback. The disconnect watcher now catches exceptions from request.receive() (e.g. a transport reset) and logs a warning instead of dying silently, making the degradation to no-op visible; a test pins that the LLM call is not cancelled in that case --------- Co-authored-by: kursad <kursad.lacin@brado.net> Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> * fix(bedrock): grant aws-external-anthropic:* in OIDC session policy for claude_platform (BerriAI#30200) (BerriAI#30205) The inline STS session policy passed to assume_role_with_web_identity acts as an IAM PERMISSION CEILING — effective permissions are the intersection of the role's identity policies and this policy. Any action not listed is silently denied even when the IAM role grants it. BerriAI#27678 added the bedrock/claude_platform/<model> route but its service-side action namespace is aws-external-anthropic:*, not bedrock:*. Without a matching statement here, every claude_platform request via OIDC (GCP federation, EKS Pod Identity webhook, etc.) 403s with 'no session policy allows the aws-external-anthropic:CreateInference action' — even with a fully permissive identity policy. Add a second ClaudePlatformLiteLLM statement covering CreateInference, CreateBatchInference, CancelBatchInference, DeleteBatchInference, CountTokens, Get*, List*. Keep aws:SecureTransport=true parity with the bedrock statement. Static creds + IRSA flow through different code paths and are not affected. Fixes BerriAI#30200 * fix(proxy): set Retry-After header on RouterRateLimitError 429 responses (BerriAI#30098) * Set Retry-After header on RouterRateLimitError responses When all deployments for a model are in cooldown, the proxy returns a 429 whose cooldown timing is only available by parsing the error message string. RouterRateLimitError already carries cooldown_time, so expose it as a standard retry-after header in _handle_llm_api_exception. The value is rounded up so clients never retry before the cooldown window ends. Fixes BerriAI#27823. * Set Retry-After after response-headers hook so cooldown wins The cooldown-derived retry-after was assigned before the post_call_response_headers_hook merge, so a callback returning a retry-after key (including a stale or empty value) silently clobbered it. Move the RouterRateLimitError block after the callback merge so the cooldown value is authoritative for this error type. * fix(router): route aspeech through async_function_with_fallbacks (BerriAI#30104) * fix(router): route aspeech through async_function_with_fallbacks Router.aspeech selected a deployment and awaited litellm.aspeech directly, so TTS requests got no retry on failure and no failover to backup deployments; the except block only fired an exception alert and re-raised. Every other router endpoint (acompletion, aembedding, atranscription, arerank) already delegates to async_function_with_fallbacks Mirror the atranscription pattern: move deployment selection and the litellm.aspeech call into a private _aspeech method, then have the public aspeech set kwargs["original_function"] = self._aspeech and await self.async_function_with_fallbacks(**kwargs). _aspeech also picks up the shared _get_async_openai_model_client helper and the same total/success/fail call accounting the sibling endpoints use Fixes BerriAI#27778. * fix(router): apply deployment kwargs and rpm semaphore in _aspeech Bring _aspeech fully in line with _atranscription: call _update_kwargs_with_deployment so deployment metadata, model_info, timeout, and default litellm params flow into the request, and wrap the litellm.aspeech call with the max_parallel_requests semaphore plus async_routing_strategy_pre_call_checks so TTS respects rpm limits the same way the other router endpoints do Also add a unit test that exercises _aspeech directly and asserts the deployment metadata reaches the underlying call * fix(slack_alerting): stop false-positive hanging request alerts for requests below the alerting threshold (BerriAI#30106) * fix(slack_alerting): skip hanging request alerts below the threshold The hanging request check alerted on any cached request whose completion status was not yet recorded, with no minimum age check. Since the background loop runs every alerting_threshold / 2 seconds, any request that happened to be in flight at a check fired a "hanging - Ns+ request time" alert even if it was only seconds old, producing a steady stream of false positives. Add a created_at timestamp to HangingRequestData, stamped when the request enters the hanging request cache, and skip requests younger than alerting_threshold without evicting them, so a later check can still alert if they never complete. Extend the cache TTL from threshold + 60s to 1.5x threshold + 60s; with the age check, entries only become alertable after threshold seconds, and the check period is threshold / 2, so the old TTL could evict a genuinely hanging request before any check saw it cross the threshold. Fixes BerriAI#27855. * fix(slack_alerting): alert once per hanging request The min-age gate stops false positives for young in-flight requests, but a genuinely hanging request still re-alerted on every checker tick within the cache TTL. With the wider TTL (1.5x threshold + 60s) that is 1-2 extra Slack notifications per stuck request at the default 600s threshold. Flag a HangingRequestData entry as alerted once its alert fires and skip flagged entries on later ticks, so each hang produces exactly one alert. The cache reference is mutated in place, so the TTL is untouched and still handles cleanup. Adds a regression test asserting one alert across multiple ticks. Fixes BerriAI#27855. * fix(health): treat all-proxy-models keys as unrestricted in /health (BerriAI#30087) * fix(health): treat all-proxy-models keys as unrestricted in /health A key granted all model permissions stores the literal "all-proxy-models" marker in its models list. The /health access filter compared that marker against real model_names, so the model list filtered down to nothing and the WebUI health check returned healthy_count=0, unhealthy_count=0 with HTTP 503. Skip the filter (both the live path and the background-cache model_id scoping) when the marker is present, matching how auth_checks treats SpecialModelNames.all_proxy_models. Fixes BerriAI#29744. * fix(health): resolve all-team-models sentinel to the team allowlist Same failure shape as the all-proxy-models case: a key carrying the literal "all-team-models" entry matches no real model_name, so the /health access filter would zero out the model list. Resolve the sentinel to the key's team models when team_id is set, matching get_key_models in model_checks.py. Without a team_id the sentinel stays unresolved and matches nothing, denying rather than widening access, mirroring _resolve_key_models_for_auth_check. * feat(proxy): auto-enable drop_params for Claude Code requests (BerriAI#30218) * feat(proxy): auto-enable drop_params for Claude Code requests Claude Code identifies itself with a claude-cli/<version> user agent and sends Anthropic-specific params (top_k, thinking, etc.) on every request. When the proxy routes those requests to a non-Anthropic provider, the unsupported params fail the call unless drop_params is configured. Detect the Claude Code user agent in add_litellm_data_to_request and default drop_params to true for those requests, without overriding an explicit drop_params value sent by the caller. * feat(proxy): respect operator litellm_settings drop_params over Claude Code default An explicit drop_params in the operator's litellm_settings (true or false) now suppresses the Claude Code user agent default, so an operator who deliberately configured drop_params: false keeps strict param validation for Claude Code clients too. The auto-default only fills the gap when neither the request body nor the config sets a value. * fix(snowflake): migrate to native endpoints with auto-routing for Claude models (BerriAI#29964) * fix(snowflake): migrate to native Cortex REST API endpoints Replaces the legacy /api/v2/cortex/inference:complete endpoint with the native OpenAI-compatible /api/v2/cortex/v1/chat/completions endpoint, fixing error 390142 (Incoming request does not contain a valid payload) when using model: snowflake/<model> in LiteLLM proxy. Changes: - litellm/llms/snowflake/chat/transformation.py: route to native /cortex/v1/chat/completions, remove Snowflake-specific tool_spec payload transformation, remove content_list response handling, add stream to supported params - litellm/llms/snowflake/anthropic/transformation.py (new): SnowflakeCortexAnthropicConfig routes Claude models to /cortex/v1/messages with anthropic-version header and Anthropic->OpenAI response transform - tests: 29 unit tests covering URL routing, auth headers, payload format, and response parsing * fix(snowflake): map max_tokens to max_completion_tokens for native endpoint * fix: handle multi-turn tool conversations and OpenAI→Anthropic tool format conversion - _extract_system_and_messages now preserves tool_calls from assistant messages and converts them to Anthropic tool_use content blocks - tool role messages are converted to user role with tool_result content blocks (as required by Anthropic Messages API) - Added _transform_tools_to_anthropic() to convert OpenAI tool format (type/function/parameters) to Anthropic format (name/input_schema) - Added comprehensive tests for multi-turn tool conversations Addresses review feedback on PR BerriAI#29964 * test: add coverage for malformed JSON and non-string tool arguments * fix(tests): update chat transformation tests for native OpenAI-compatible endpoint * style: apply black formatting * fix: resolve mypy type errors in anthropic transformation * fix: correct mypy type: ignore error codes (attr-defined) * fix: use max_tokens instead of max_completion_tokens for Snowflake endpoint compatibility * refactor: merge Anthropic config into unified SnowflakeConfig with auto-routing - Remove separate SnowflakeCortexAnthropicConfig and anthropic/ directory - SnowflakeConfig now auto-routes based on model name: - Claude models → /messages endpoint (Anthropic format) - All others → /chat/completions endpoint (OpenAI format) - No new provider needed (stays as SNOWFLAKE = 'snowflake') - Tool message transformation for Claude: tool_calls → tool_use blocks, tool role → user with tool_result - OpenAI → Anthropic tool format conversion (parameters → input_schema) - Addresses Greptile feedback about unwired SnowflakeCortexAnthropicConfig * fix: use max_completion_tokens for /chat/completions (Snowflake deprecated max_tokens on this endpoint) * fix(tests): update assertions for Claude auto-routing to /messages endpoint * fix(snowflake): add tool_choice conversion and preserve max_completion_tokens in Anthropic path * fix(snowflake): use ChatCompletionMessageToolCall objects and strip model prefix on OpenAI path * fix(snowflake): collect multiple system messages to prevent guardrail override * chore: remove committed .pyc files and add __pycache__ to .gitignore * fix: remove unused Union import * fix: restore original .gitignore (accidentally replaced in earlier commit) * feat(snowflake): add streaming response handler for both Anthropic and OpenAI SSE formats * fix: remove unused AsyncIterator and Iterator imports * fix: add missing total_tokens to ChatCompletionUsageBlock * fix(snowflake): coalesce consecutive tool results into single user message for Anthropic * fix(snowflake): handle message_start event for streaming input_tokens tracking * fix: evict last deleted model in multi-instance deployments (BerriAI#28608) * fix: evict last deleted model in multi-instance deployments _delete_deployment had an early return when db_models was empty, preventing eviction of the last deleted model during reconciliation. - Remove len(db_models)==0 early return from _delete_deployment - Return None (not []) from _get_models_from_db on DB failure so callers can distinguish a transient failure from a genuinely empty DB - Guard _update_llm_router against None to skip updates on DB failure Fixes BerriAI#28443 * test: remove dead MagicMock assignment in type_mismatch test * fix: update test to pass [] not None to _update_llm_router test_ProxyConfig__update_llm_router_bad_proxy_logging_raises was passing None as new_models to get through to the proxy_logging_obj check, but the None guard we added now returns early before reaching that path. Pass [] instead so the test exercises the intended AttributeError case. Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> * chore: regenerate API types to sync schema.d.ts with proxy OpenAPI spec Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> --------- Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> * fix: invalidate Redis spend counter on /key/reset_spend (BerriAI#29694) * fix: set Redis spend counter to reset_to value on /key/reset_spend Previously, the Redis spend counter was always set to 0.0 after a reset, even when reset_to was a non-zero value (partial reset). This caused the budget to be under-enforced for up to 60 seconds until the counter expired and fell through to the DB. Now the counter is set to the actual reset_to value, so partial resets are reflected correctly and budget enforcement is consistent. * test: update reset_key_spend test to match direct cache set The implementation now sets spend_counter_cache directly instead of calling _invalidate_spend_counter. Update the test to verify the in_memory_cache.set_cache call with the correct key, value, and ttl. --------- Co-authored-by: michaelxer <michaelxer@users.noreply.github.com> * fix: add scaleway models pricing (BerriAI#27659) * fix: Add embeddings support for Scaleway provider * fix: resolve merge conflicts * fix(main): clarify backend route handling for Swagger static assets (BerriAI#30196) * fix(main): clarify backend route handling for Swagger static assets * fix(allowlist): add BACKEND_MOUNT_PATHS for Swagger static assets * fix(voyage): route multimodal embeddings to correct endpoint (BerriAI#30193) * fix(voyage): route multimodal embeddings to correct endpoint * test(voyage): cover multimodal embedding edge cases * test(voyage): cover api key fallback * fix(voyage): raise early on missing api key and malformed image url * test(voyage): cover utils routing and helper * fix(voyage): route supported openai params for multimodal models * style: apply black formatting * fix(ui): infer Azure API version from API base (BerriAI#30204) * fix(ui): infer Azure API version from API base * fix(ui): address Azure API version feedback * Update litellm/llms/snowflake/chat/transformation.py Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> * feat(datadog): add team-scoped Datadog callback support (BerriAI#29947) Enable teams to configure their own Datadog credentials via POST /team/{team_id}/callback, following the same pattern as Langfuse. * Merge pull request BerriAI#29528 from aanchal22/litellm_byok-alias-merge fix(proxy): atomic merge for team model aliases and team.models on BYOK create * feat: add EmpirioLabs as an OpenAI-compatible provider (BerriAI#30278) Co-authored-by: Adam Dalloul <adam.d.developer@gmail.com> * fix: resolve failing tests and lint in snowflake/team endpoints - Black-format snowflake/chat/transformation.py to fix lint failure - Update Anthropic config test to expect default max_tokens of 4096 (matches implementation) - Add AsyncMock + execute_raw mock to team_model_add cache-refresh pin test - Add model_dump mock and patch cache/logging in test_uses_atomic_array_append_with_dedup Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * fix(test): update test_db_error_new_model_check for new _delete_deployment logic _delete_deployment no longer short-circuits on empty db_models — it now treats [] as a valid empty-DB state and proceeds to check config models. Mock get_config to return the two router deployments so they appear in combined_id_list and are protected, which matches the real-world scenario where a DB error occurs but the models are config-backed. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * feat(proxy): register cancel_on_disconnect in ConfigGeneralSettings and config list (BerriAI#30295) * feat(proxy): register cancel_on_disconnect in ConfigGeneralSettings and config list Follow-up to BerriAI#30223 per maintainer review: documents the flag in ConfigGeneralSettings with a short description and adds it to allowed_args in get_config_list so the UI and /config/list expose it. A test pins that /config/list returns the field with type Boolean, which requires both registrations to be present * chore(ui): regenerate schema.d.ts for cancel_on_disconnect --------- Co-authored-by: kursad <kursad.lacin@brado.net> * fix(datadog): never fall back to env DD_API_KEY for caller-supplied destinations Team/key-scoped Datadog loggers could be pointed at an arbitrary dd_agent_host or dd_site while omitting dd_api_key, causing the proxy's global DD_API_KEY to be sent as the DD-API-KEY header to that destination. Gate the env-var fallback behind an allow_env_credentials flag, set to False when the destination is caller-supplied, mirroring the existing langfuse/langsmith pattern. --------- Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> Co-authored-by: Emerson Gomes <emerson.gomes@thalesgroup.com> Co-authored-by: daitran-tensormesh <dai@tensormesh.ai> Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> Co-authored-by: Muspi Merol <me@promplate.dev> Co-authored-by: fangkang <fangkangm@gmail.com> Co-authored-by: Chris Hoogeboom <chris.hoogeboom@gmail.com> Co-authored-by: kursadlacin <kursadlacin@gmail.com> Co-authored-by: kursad <kursad.lacin@brado.net> Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> Co-authored-by: hcl <chenglunhu@gmail.com> Co-authored-by: Filippo Menghi <113345637+Cyberfilo@users.noreply.github.com> Co-authored-by: Mateo Wang <277851410+mateo-berri@users.noreply.github.com> Co-authored-by: sfc-gh-nashukla <navnit.shukla@snowflake.com> Co-authored-by: Rudra Dudhat <contact.rdudhat@gmail.com> Co-authored-by: Michael <52305679+michaelxer@users.noreply.github.com> Co-authored-by: michaelxer <michaelxer@users.noreply.github.com> Co-authored-by: Quentin Champenois <26109239+Quentinchampenois@users.noreply.github.com> Co-authored-by: mauriceberentsen <mauriceberentsen@live.nl> Co-authored-by: lost9999 <56498264+lost9999@users.noreply.github.com> Co-authored-by: GaetanVDB07 <86427581+GaetanVDB07@users.noreply.github.com> Co-authored-by: Aanchal Khandelwal <aan2210khandelwal@gmail.com> Co-authored-by: Adam Dalloul <adam_dalloul@icloud.com> Co-authored-by: Adam Dalloul <adam.d.developer@gmail.com> Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
factnn
pushed a commit
to factnn/litellm
that referenced
this pull request
Jun 18, 2026
* feat(bedrock): add bedrock mantle gemma 4 models (BerriAI#30264) * feat(bedrock): add bedrock mantle gemma 4 models * test(bedrock): harden mantle local cost fixture * feat(responses): enable the responses API for the Tensormesh provider (BerriAI#30209) * feat(responses): enable the responses API for the Tensormesh provider * Update litellm/llms/openai_like/providers.json Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> --------- Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> * fix(langfuse_otel): mark LLM spans as generations (BerriAI#30250) * fix(bedrock): stop stream_chunk_size leaking into invoke request bodies (BerriAI#30240) stream_chunk_size is a LiteLLM-internal knob for re-chunking the HTTP response stream. The invoke transformations splat optional_params into the provider request body without dropping it, and Bedrock rejects unknown fields, so any bedrock/invoke request that sets the parameter fails with ValidationException: stream_chunk_size: Extra inputs are not permitted. Drop it in the invoke dispatcher (covers cohere, titan, mistral, meta, ai21) and in the Claude messages-format request builder (the route used for bedrock/invoke Anthropic models) * fix(bedrock): stop buffering streamed tool-call argument deltas (BerriAI#30231) * fix(bedrock): stop buffering streamed tool-call argument deltas Two issues made Bedrock tool-use streaming arrive as a single end-of-stream burst through LiteLLM while plain text streamed fine. First, the anthropic-beta allowlist mapped fine-grained-tool-streaming-2025-05-14 to null for bedrock and bedrock_converse, so the header was silently stripped. Without that beta, Anthropic models on Bedrock buffer tool input server-side and emit all toolUse.input deltas at once (verified against converse-stream and invoke-with-response-stream directly). Bedrock accepts the beta via additionalModelRequestFields.anthropic_beta, so it is now forwarded. Second, the streaming reads re-chunked the AWS event stream with iter_bytes(chunk_size=1024). httpx's ByteChunker only releases full 1024-byte blocks, so the small early events (messageStart, contentBlockStart, first deltas) sat in the buffer until enough bytes accumulated, pushing time-to-first-byte from ~1.4s to ~8.5s on buffered tool-use streams. The default is now no re-chunking; an explicit stream_chunk_size is still honored. * test(bedrock): cover explicit stream_chunk_size on sync invoke path * test(bedrock): cover stream_chunk_size plumbing through converse completion * test(bedrock): cover stream_chunk_size default in legacy BedrockLLM streaming * test(bedrock): merge converse handler tests into existing mapped test file pytest imports test modules by basename in non-package test dirs, so the new tests/test_litellm/llms/bedrock/chat/test_converse_handler.py collided with the pre-existing tests/test_litellm/llms/chat/test_converse_handler.py and broke collection in CI. Move the new tests into the existing file * feat(otel): emit v2 cost breakdown + stamp tracer scope version (BerriAI#30156) Read the StandardLoggingPayload cost_breakdown into a typed LLMCost on LLMCallSpanData and emit each component under litellm.cost.* (absent components omitted, so spans stay sparse). Stamp litellm.__version__ as the instrumentation scope version so every v2 span carries a deterministic scope.version. Tests under tests/test_litellm/integrations/otel/. * fix(proxy): cancel in-flight upstream LLM request on client disconnect (opt-in) (BerriAI#30223) * fix(proxy): cancel in-flight upstream LLM request on client disconnect (opt-in) On the non-streaming path, base_process_llm_request awaited the LLM call with no disconnect monitoring; when the HTTP client went away the upstream request kept running until completion or request_timeout (6000s default), holding a backend slot (e.g. a vLLM GPU slot) for output nobody would read Add an opt-in general_settings.cancel_on_disconnect flag, default off, so the default code path is unchanged. When enabled, a receive-based watcher task observes http.disconnect and cancels the asyncio.gather driving the upstream call. The resulting CancelledError is converted to HTTPException 499 only when the disconnect event is set, so server-initiated cancellations still propagate as-is. The 499 then flows through _handle_llm_api_exception like any other failure, meaning post_call_failure_hook still releases max_parallel_requests slots and fires spend and alerting callbacks; it is logged at info level instead of a full traceback Also removes the dead check_request_disconnection helper in proxy_server.py (zero call sites) along with its behavior-pin tests Builds on the receive-based design from BerriAI#25776 Addresses BerriAI#13774. Re-fixes BerriAI#22805 (regressed after the BerriAI#14295 revert) Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> * fix(proxy): scope 499 quiet logging to disconnects and harden watcher Address the two P2 findings from the Greptile review on BerriAI#30223. The info-level logging in _log_llm_api_exception now applies only to the disconnect-specific HTTPException (status 499 plus the shared _CLIENT_DISCONNECT_DETAIL message), so any other 499 raised by hooks or guardrails keeps its full traceback. The disconnect watcher now catches exceptions from request.receive() (e.g. a transport reset) and logs a warning instead of dying silently, making the degradation to no-op visible; a test pins that the LLM call is not cancelled in that case --------- Co-authored-by: kursad <kursad.lacin@brado.net> Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> * fix(bedrock): grant aws-external-anthropic:* in OIDC session policy for claude_platform (BerriAI#30200) (BerriAI#30205) The inline STS session policy passed to assume_role_with_web_identity acts as an IAM PERMISSION CEILING — effective permissions are the intersection of the role's identity policies and this policy. Any action not listed is silently denied even when the IAM role grants it. BerriAI#27678 added the bedrock/claude_platform/<model> route but its service-side action namespace is aws-external-anthropic:*, not bedrock:*. Without a matching statement here, every claude_platform request via OIDC (GCP federation, EKS Pod Identity webhook, etc.) 403s with 'no session policy allows the aws-external-anthropic:CreateInference action' — even with a fully permissive identity policy. Add a second ClaudePlatformLiteLLM statement covering CreateInference, CreateBatchInference, CancelBatchInference, DeleteBatchInference, CountTokens, Get*, List*. Keep aws:SecureTransport=true parity with the bedrock statement. Static creds + IRSA flow through different code paths and are not affected. Fixes BerriAI#30200 * fix(proxy): set Retry-After header on RouterRateLimitError 429 responses (BerriAI#30098) * Set Retry-After header on RouterRateLimitError responses When all deployments for a model are in cooldown, the proxy returns a 429 whose cooldown timing is only available by parsing the error message string. RouterRateLimitError already carries cooldown_time, so expose it as a standard retry-after header in _handle_llm_api_exception. The value is rounded up so clients never retry before the cooldown window ends. Fixes BerriAI#27823. * Set Retry-After after response-headers hook so cooldown wins The cooldown-derived retry-after was assigned before the post_call_response_headers_hook merge, so a callback returning a retry-after key (including a stale or empty value) silently clobbered it. Move the RouterRateLimitError block after the callback merge so the cooldown value is authoritative for this error type. * fix(router): route aspeech through async_function_with_fallbacks (BerriAI#30104) * fix(router): route aspeech through async_function_with_fallbacks Router.aspeech selected a deployment and awaited litellm.aspeech directly, so TTS requests got no retry on failure and no failover to backup deployments; the except block only fired an exception alert and re-raised. Every other router endpoint (acompletion, aembedding, atranscription, arerank) already delegates to async_function_with_fallbacks Mirror the atranscription pattern: move deployment selection and the litellm.aspeech call into a private _aspeech method, then have the public aspeech set kwargs["original_function"] = self._aspeech and await self.async_function_with_fallbacks(**kwargs). _aspeech also picks up the shared _get_async_openai_model_client helper and the same total/success/fail call accounting the sibling endpoints use Fixes BerriAI#27778. * fix(router): apply deployment kwargs and rpm semaphore in _aspeech Bring _aspeech fully in line with _atranscription: call _update_kwargs_with_deployment so deployment metadata, model_info, timeout, and default litellm params flow into the request, and wrap the litellm.aspeech call with the max_parallel_requests semaphore plus async_routing_strategy_pre_call_checks so TTS respects rpm limits the same way the other router endpoints do Also add a unit test that exercises _aspeech directly and asserts the deployment metadata reaches the underlying call * fix(slack_alerting): stop false-positive hanging request alerts for requests below the alerting threshold (BerriAI#30106) * fix(slack_alerting): skip hanging request alerts below the threshold The hanging request check alerted on any cached request whose completion status was not yet recorded, with no minimum age check. Since the background loop runs every alerting_threshold / 2 seconds, any request that happened to be in flight at a check fired a "hanging - Ns+ request time" alert even if it was only seconds old, producing a steady stream of false positives. Add a created_at timestamp to HangingRequestData, stamped when the request enters the hanging request cache, and skip requests younger than alerting_threshold without evicting them, so a later check can still alert if they never complete. Extend the cache TTL from threshold + 60s to 1.5x threshold + 60s; with the age check, entries only become alertable after threshold seconds, and the check period is threshold / 2, so the old TTL could evict a genuinely hanging request before any check saw it cross the threshold. Fixes BerriAI#27855. * fix(slack_alerting): alert once per hanging request The min-age gate stops false positives for young in-flight requests, but a genuinely hanging request still re-alerted on every checker tick within the cache TTL. With the wider TTL (1.5x threshold + 60s) that is 1-2 extra Slack notifications per stuck request at the default 600s threshold. Flag a HangingRequestData entry as alerted once its alert fires and skip flagged entries on later ticks, so each hang produces exactly one alert. The cache reference is mutated in place, so the TTL is untouched and still handles cleanup. Adds a regression test asserting one alert across multiple ticks. Fixes BerriAI#27855. * fix(health): treat all-proxy-models keys as unrestricted in /health (BerriAI#30087) * fix(health): treat all-proxy-models keys as unrestricted in /health A key granted all model permissions stores the literal "all-proxy-models" marker in its models list. The /health access filter compared that marker against real model_names, so the model list filtered down to nothing and the WebUI health check returned healthy_count=0, unhealthy_count=0 with HTTP 503. Skip the filter (both the live path and the background-cache model_id scoping) when the marker is present, matching how auth_checks treats SpecialModelNames.all_proxy_models. Fixes BerriAI#29744. * fix(health): resolve all-team-models sentinel to the team allowlist Same failure shape as the all-proxy-models case: a key carrying the literal "all-team-models" entry matches no real model_name, so the /health access filter would zero out the model list. Resolve the sentinel to the key's team models when team_id is set, matching get_key_models in model_checks.py. Without a team_id the sentinel stays unresolved and matches nothing, denying rather than widening access, mirroring _resolve_key_models_for_auth_check. * feat(proxy): auto-enable drop_params for Claude Code requests (BerriAI#30218) * feat(proxy): auto-enable drop_params for Claude Code requests Claude Code identifies itself with a claude-cli/<version> user agent and sends Anthropic-specific params (top_k, thinking, etc.) on every request. When the proxy routes those requests to a non-Anthropic provider, the unsupported params fail the call unless drop_params is configured. Detect the Claude Code user agent in add_litellm_data_to_request and default drop_params to true for those requests, without overriding an explicit drop_params value sent by the caller. * feat(proxy): respect operator litellm_settings drop_params over Claude Code default An explicit drop_params in the operator's litellm_settings (true or false) now suppresses the Claude Code user agent default, so an operator who deliberately configured drop_params: false keeps strict param validation for Claude Code clients too. The auto-default only fills the gap when neither the request body nor the config sets a value. * fix(snowflake): migrate to native endpoints with auto-routing for Claude models (BerriAI#29964) * fix(snowflake): migrate to native Cortex REST API endpoints Replaces the legacy /api/v2/cortex/inference:complete endpoint with the native OpenAI-compatible /api/v2/cortex/v1/chat/completions endpoint, fixing error 390142 (Incoming request does not contain a valid payload) when using model: snowflake/<model> in LiteLLM proxy. Changes: - litellm/llms/snowflake/chat/transformation.py: route to native /cortex/v1/chat/completions, remove Snowflake-specific tool_spec payload transformation, remove content_list response handling, add stream to supported params - litellm/llms/snowflake/anthropic/transformation.py (new): SnowflakeCortexAnthropicConfig routes Claude models to /cortex/v1/messages with anthropic-version header and Anthropic->OpenAI response transform - tests: 29 unit tests covering URL routing, auth headers, payload format, and response parsing * fix(snowflake): map max_tokens to max_completion_tokens for native endpoint * fix: handle multi-turn tool conversations and OpenAI→Anthropic tool format conversion - _extract_system_and_messages now preserves tool_calls from assistant messages and converts them to Anthropic tool_use content blocks - tool role messages are converted to user role with tool_result content blocks (as required by Anthropic Messages API) - Added _transform_tools_to_anthropic() to convert OpenAI tool format (type/function/parameters) to Anthropic format (name/input_schema) - Added comprehensive tests for multi-turn tool conversations Addresses review feedback on PR BerriAI#29964 * test: add coverage for malformed JSON and non-string tool arguments * fix(tests): update chat transformation tests for native OpenAI-compatible endpoint * style: apply black formatting * fix: resolve mypy type errors in anthropic transformation * fix: correct mypy type: ignore error codes (attr-defined) * fix: use max_tokens instead of max_completion_tokens for Snowflake endpoint compatibility * refactor: merge Anthropic config into unified SnowflakeConfig with auto-routing - Remove separate SnowflakeCortexAnthropicConfig and anthropic/ directory - SnowflakeConfig now auto-routes based on model name: - Claude models → /messages endpoint (Anthropic format) - All others → /chat/completions endpoint (OpenAI format) - No new provider needed (stays as SNOWFLAKE = 'snowflake') - Tool message transformation for Claude: tool_calls → tool_use blocks, tool role → user with tool_result - OpenAI → Anthropic tool format conversion (parameters → input_schema) - Addresses Greptile feedback about unwired SnowflakeCortexAnthropicConfig * fix: use max_completion_tokens for /chat/completions (Snowflake deprecated max_tokens on this endpoint) * fix(tests): update assertions for Claude auto-routing to /messages endpoint * fix(snowflake): add tool_choice conversion and preserve max_completion_tokens in Anthropic path * fix(snowflake): use ChatCompletionMessageToolCall objects and strip model prefix on OpenAI path * fix(snowflake): collect multiple system messages to prevent guardrail override * chore: remove committed .pyc files and add __pycache__ to .gitignore * fix: remove unused Union import * fix: restore original .gitignore (accidentally replaced in earlier commit) * feat(snowflake): add streaming response handler for both Anthropic and OpenAI SSE formats * fix: remove unused AsyncIterator and Iterator imports * fix: add missing total_tokens to ChatCompletionUsageBlock * fix(snowflake): coalesce consecutive tool results into single user message for Anthropic * fix(snowflake): handle message_start event for streaming input_tokens tracking * fix: evict last deleted model in multi-instance deployments (BerriAI#28608) * fix: evict last deleted model in multi-instance deployments _delete_deployment had an early return when db_models was empty, preventing eviction of the last deleted model during reconciliation. - Remove len(db_models)==0 early return from _delete_deployment - Return None (not []) from _get_models_from_db on DB failure so callers can distinguish a transient failure from a genuinely empty DB - Guard _update_llm_router against None to skip updates on DB failure Fixes BerriAI#28443 * test: remove dead MagicMock assignment in type_mismatch test * fix: update test to pass [] not None to _update_llm_router test_ProxyConfig__update_llm_router_bad_proxy_logging_raises was passing None as new_models to get through to the proxy_logging_obj check, but the None guard we added now returns early before reaching that path. Pass [] instead so the test exercises the intended AttributeError case. Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> * chore: regenerate API types to sync schema.d.ts with proxy OpenAPI spec Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> --------- Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> * fix: invalidate Redis spend counter on /key/reset_spend (BerriAI#29694) * fix: set Redis spend counter to reset_to value on /key/reset_spend Previously, the Redis spend counter was always set to 0.0 after a reset, even when reset_to was a non-zero value (partial reset). This caused the budget to be under-enforced for up to 60 seconds until the counter expired and fell through to the DB. Now the counter is set to the actual reset_to value, so partial resets are reflected correctly and budget enforcement is consistent. * test: update reset_key_spend test to match direct cache set The implementation now sets spend_counter_cache directly instead of calling _invalidate_spend_counter. Update the test to verify the in_memory_cache.set_cache call with the correct key, value, and ttl. --------- Co-authored-by: michaelxer <michaelxer@users.noreply.github.com> * fix: add scaleway models pricing (BerriAI#27659) * fix: Add embeddings support for Scaleway provider * fix: resolve merge conflicts * fix(main): clarify backend route handling for Swagger static assets (BerriAI#30196) * fix(main): clarify backend route handling for Swagger static assets * fix(allowlist): add BACKEND_MOUNT_PATHS for Swagger static assets * fix(voyage): route multimodal embeddings to correct endpoint (BerriAI#30193) * fix(voyage): route multimodal embeddings to correct endpoint * test(voyage): cover multimodal embedding edge cases * test(voyage): cover api key fallback * fix(voyage): raise early on missing api key and malformed image url * test(voyage): cover utils routing and helper * fix(voyage): route supported openai params for multimodal models * style: apply black formatting * fix(ui): infer Azure API version from API base (BerriAI#30204) * fix(ui): infer Azure API version from API base * fix(ui): address Azure API version feedback * Update litellm/llms/snowflake/chat/transformation.py Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> * feat(datadog): add team-scoped Datadog callback support (BerriAI#29947) Enable teams to configure their own Datadog credentials via POST /team/{team_id}/callback, following the same pattern as Langfuse. * Merge pull request BerriAI#29528 from aanchal22/litellm_byok-alias-merge fix(proxy): atomic merge for team model aliases and team.models on BYOK create * feat: add EmpirioLabs as an OpenAI-compatible provider (BerriAI#30278) Co-authored-by: Adam Dalloul <adam.d.developer@gmail.com> * fix: resolve failing tests and lint in snowflake/team endpoints - Black-format snowflake/chat/transformation.py to fix lint failure - Update Anthropic config test to expect default max_tokens of 4096 (matches implementation) - Add AsyncMock + execute_raw mock to team_model_add cache-refresh pin test - Add model_dump mock and patch cache/logging in test_uses_atomic_array_append_with_dedup Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * fix(test): update test_db_error_new_model_check for new _delete_deployment logic _delete_deployment no longer short-circuits on empty db_models — it now treats [] as a valid empty-DB state and proceeds to check config models. Mock get_config to return the two router deployments so they appear in combined_id_list and are protected, which matches the real-world scenario where a DB error occurs but the models are config-backed. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * feat(proxy): register cancel_on_disconnect in ConfigGeneralSettings and config list (BerriAI#30295) * feat(proxy): register cancel_on_disconnect in ConfigGeneralSettings and config list Follow-up to BerriAI#30223 per maintainer review: documents the flag in ConfigGeneralSettings with a short description and adds it to allowed_args in get_config_list so the UI and /config/list expose it. A test pins that /config/list returns the field with type Boolean, which requires both registrations to be present * chore(ui): regenerate schema.d.ts for cancel_on_disconnect --------- Co-authored-by: kursad <kursad.lacin@brado.net> * fix(datadog): never fall back to env DD_API_KEY for caller-supplied destinations Team/key-scoped Datadog loggers could be pointed at an arbitrary dd_agent_host or dd_site while omitting dd_api_key, causing the proxy's global DD_API_KEY to be sent as the DD-API-KEY header to that destination. Gate the env-var fallback behind an allow_env_credentials flag, set to False when the destination is caller-supplied, mirroring the existing langfuse/langsmith pattern. --------- Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> Co-authored-by: Emerson Gomes <emerson.gomes@thalesgroup.com> Co-authored-by: daitran-tensormesh <dai@tensormesh.ai> Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> Co-authored-by: Muspi Merol <me@promplate.dev> Co-authored-by: fangkang <fangkangm@gmail.com> Co-authored-by: Chris Hoogeboom <chris.hoogeboom@gmail.com> Co-authored-by: kursadlacin <kursadlacin@gmail.com> Co-authored-by: kursad <kursad.lacin@brado.net> Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> Co-authored-by: hcl <chenglunhu@gmail.com> Co-authored-by: Filippo Menghi <113345637+Cyberfilo@users.noreply.github.com> Co-authored-by: Mateo Wang <277851410+mateo-berri@users.noreply.github.com> Co-authored-by: sfc-gh-nashukla <navnit.shukla@snowflake.com> Co-authored-by: Rudra Dudhat <contact.rdudhat@gmail.com> Co-authored-by: Michael <52305679+michaelxer@users.noreply.github.com> Co-authored-by: michaelxer <michaelxer@users.noreply.github.com> Co-authored-by: Quentin Champenois <26109239+Quentinchampenois@users.noreply.github.com> Co-authored-by: mauriceberentsen <mauriceberentsen@live.nl> Co-authored-by: lost9999 <56498264+lost9999@users.noreply.github.com> Co-authored-by: GaetanVDB07 <86427581+GaetanVDB07@users.noreply.github.com> Co-authored-by: Aanchal Khandelwal <aan2210khandelwal@gmail.com> Co-authored-by: Adam Dalloul <adam_dalloul@icloud.com> Co-authored-by: Adam Dalloul <adam.d.developer@gmail.com> Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
fzowl
pushed a commit
to fzowl/litellm
that referenced
this pull request
Jun 24, 2026
* feat(bedrock): add bedrock mantle gemma 4 models (BerriAI#30264) * feat(bedrock): add bedrock mantle gemma 4 models * test(bedrock): harden mantle local cost fixture * feat(responses): enable the responses API for the Tensormesh provider (BerriAI#30209) * feat(responses): enable the responses API for the Tensormesh provider * Update litellm/llms/openai_like/providers.json Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> --------- Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> * fix(langfuse_otel): mark LLM spans as generations (BerriAI#30250) * fix(bedrock): stop stream_chunk_size leaking into invoke request bodies (BerriAI#30240) stream_chunk_size is a LiteLLM-internal knob for re-chunking the HTTP response stream. The invoke transformations splat optional_params into the provider request body without dropping it, and Bedrock rejects unknown fields, so any bedrock/invoke request that sets the parameter fails with ValidationException: stream_chunk_size: Extra inputs are not permitted. Drop it in the invoke dispatcher (covers cohere, titan, mistral, meta, ai21) and in the Claude messages-format request builder (the route used for bedrock/invoke Anthropic models) * fix(bedrock): stop buffering streamed tool-call argument deltas (BerriAI#30231) * fix(bedrock): stop buffering streamed tool-call argument deltas Two issues made Bedrock tool-use streaming arrive as a single end-of-stream burst through LiteLLM while plain text streamed fine. First, the anthropic-beta allowlist mapped fine-grained-tool-streaming-2025-05-14 to null for bedrock and bedrock_converse, so the header was silently stripped. Without that beta, Anthropic models on Bedrock buffer tool input server-side and emit all toolUse.input deltas at once (verified against converse-stream and invoke-with-response-stream directly). Bedrock accepts the beta via additionalModelRequestFields.anthropic_beta, so it is now forwarded. Second, the streaming reads re-chunked the AWS event stream with iter_bytes(chunk_size=1024). httpx's ByteChunker only releases full 1024-byte blocks, so the small early events (messageStart, contentBlockStart, first deltas) sat in the buffer until enough bytes accumulated, pushing time-to-first-byte from ~1.4s to ~8.5s on buffered tool-use streams. The default is now no re-chunking; an explicit stream_chunk_size is still honored. * test(bedrock): cover explicit stream_chunk_size on sync invoke path * test(bedrock): cover stream_chunk_size plumbing through converse completion * test(bedrock): cover stream_chunk_size default in legacy BedrockLLM streaming * test(bedrock): merge converse handler tests into existing mapped test file pytest imports test modules by basename in non-package test dirs, so the new tests/test_litellm/llms/bedrock/chat/test_converse_handler.py collided with the pre-existing tests/test_litellm/llms/chat/test_converse_handler.py and broke collection in CI. Move the new tests into the existing file * feat(otel): emit v2 cost breakdown + stamp tracer scope version (BerriAI#30156) Read the StandardLoggingPayload cost_breakdown into a typed LLMCost on LLMCallSpanData and emit each component under litellm.cost.* (absent components omitted, so spans stay sparse). Stamp litellm.__version__ as the instrumentation scope version so every v2 span carries a deterministic scope.version. Tests under tests/test_litellm/integrations/otel/. * fix(proxy): cancel in-flight upstream LLM request on client disconnect (opt-in) (BerriAI#30223) * fix(proxy): cancel in-flight upstream LLM request on client disconnect (opt-in) On the non-streaming path, base_process_llm_request awaited the LLM call with no disconnect monitoring; when the HTTP client went away the upstream request kept running until completion or request_timeout (6000s default), holding a backend slot (e.g. a vLLM GPU slot) for output nobody would read Add an opt-in general_settings.cancel_on_disconnect flag, default off, so the default code path is unchanged. When enabled, a receive-based watcher task observes http.disconnect and cancels the asyncio.gather driving the upstream call. The resulting CancelledError is converted to HTTPException 499 only when the disconnect event is set, so server-initiated cancellations still propagate as-is. The 499 then flows through _handle_llm_api_exception like any other failure, meaning post_call_failure_hook still releases max_parallel_requests slots and fires spend and alerting callbacks; it is logged at info level instead of a full traceback Also removes the dead check_request_disconnection helper in proxy_server.py (zero call sites) along with its behavior-pin tests Builds on the receive-based design from BerriAI#25776 Addresses BerriAI#13774. Re-fixes BerriAI#22805 (regressed after the BerriAI#14295 revert) Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> * fix(proxy): scope 499 quiet logging to disconnects and harden watcher Address the two P2 findings from the Greptile review on BerriAI#30223. The info-level logging in _log_llm_api_exception now applies only to the disconnect-specific HTTPException (status 499 plus the shared _CLIENT_DISCONNECT_DETAIL message), so any other 499 raised by hooks or guardrails keeps its full traceback. The disconnect watcher now catches exceptions from request.receive() (e.g. a transport reset) and logs a warning instead of dying silently, making the degradation to no-op visible; a test pins that the LLM call is not cancelled in that case --------- Co-authored-by: kursad <kursad.lacin@brado.net> Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> * fix(bedrock): grant aws-external-anthropic:* in OIDC session policy for claude_platform (BerriAI#30200) (BerriAI#30205) The inline STS session policy passed to assume_role_with_web_identity acts as an IAM PERMISSION CEILING — effective permissions are the intersection of the role's identity policies and this policy. Any action not listed is silently denied even when the IAM role grants it. BerriAI#27678 added the bedrock/claude_platform/<model> route but its service-side action namespace is aws-external-anthropic:*, not bedrock:*. Without a matching statement here, every claude_platform request via OIDC (GCP federation, EKS Pod Identity webhook, etc.) 403s with 'no session policy allows the aws-external-anthropic:CreateInference action' — even with a fully permissive identity policy. Add a second ClaudePlatformLiteLLM statement covering CreateInference, CreateBatchInference, CancelBatchInference, DeleteBatchInference, CountTokens, Get*, List*. Keep aws:SecureTransport=true parity with the bedrock statement. Static creds + IRSA flow through different code paths and are not affected. Fixes BerriAI#30200 * fix(proxy): set Retry-After header on RouterRateLimitError 429 responses (BerriAI#30098) * Set Retry-After header on RouterRateLimitError responses When all deployments for a model are in cooldown, the proxy returns a 429 whose cooldown timing is only available by parsing the error message string. RouterRateLimitError already carries cooldown_time, so expose it as a standard retry-after header in _handle_llm_api_exception. The value is rounded up so clients never retry before the cooldown window ends. Fixes BerriAI#27823. * Set Retry-After after response-headers hook so cooldown wins The cooldown-derived retry-after was assigned before the post_call_response_headers_hook merge, so a callback returning a retry-after key (including a stale or empty value) silently clobbered it. Move the RouterRateLimitError block after the callback merge so the cooldown value is authoritative for this error type. * fix(router): route aspeech through async_function_with_fallbacks (BerriAI#30104) * fix(router): route aspeech through async_function_with_fallbacks Router.aspeech selected a deployment and awaited litellm.aspeech directly, so TTS requests got no retry on failure and no failover to backup deployments; the except block only fired an exception alert and re-raised. Every other router endpoint (acompletion, aembedding, atranscription, arerank) already delegates to async_function_with_fallbacks Mirror the atranscription pattern: move deployment selection and the litellm.aspeech call into a private _aspeech method, then have the public aspeech set kwargs["original_function"] = self._aspeech and await self.async_function_with_fallbacks(**kwargs). _aspeech also picks up the shared _get_async_openai_model_client helper and the same total/success/fail call accounting the sibling endpoints use Fixes BerriAI#27778. * fix(router): apply deployment kwargs and rpm semaphore in _aspeech Bring _aspeech fully in line with _atranscription: call _update_kwargs_with_deployment so deployment metadata, model_info, timeout, and default litellm params flow into the request, and wrap the litellm.aspeech call with the max_parallel_requests semaphore plus async_routing_strategy_pre_call_checks so TTS respects rpm limits the same way the other router endpoints do Also add a unit test that exercises _aspeech directly and asserts the deployment metadata reaches the underlying call * fix(slack_alerting): stop false-positive hanging request alerts for requests below the alerting threshold (BerriAI#30106) * fix(slack_alerting): skip hanging request alerts below the threshold The hanging request check alerted on any cached request whose completion status was not yet recorded, with no minimum age check. Since the background loop runs every alerting_threshold / 2 seconds, any request that happened to be in flight at a check fired a "hanging - Ns+ request time" alert even if it was only seconds old, producing a steady stream of false positives. Add a created_at timestamp to HangingRequestData, stamped when the request enters the hanging request cache, and skip requests younger than alerting_threshold without evicting them, so a later check can still alert if they never complete. Extend the cache TTL from threshold + 60s to 1.5x threshold + 60s; with the age check, entries only become alertable after threshold seconds, and the check period is threshold / 2, so the old TTL could evict a genuinely hanging request before any check saw it cross the threshold. Fixes BerriAI#27855. * fix(slack_alerting): alert once per hanging request The min-age gate stops false positives for young in-flight requests, but a genuinely hanging request still re-alerted on every checker tick within the cache TTL. With the wider TTL (1.5x threshold + 60s) that is 1-2 extra Slack notifications per stuck request at the default 600s threshold. Flag a HangingRequestData entry as alerted once its alert fires and skip flagged entries on later ticks, so each hang produces exactly one alert. The cache reference is mutated in place, so the TTL is untouched and still handles cleanup. Adds a regression test asserting one alert across multiple ticks. Fixes BerriAI#27855. * fix(health): treat all-proxy-models keys as unrestricted in /health (BerriAI#30087) * fix(health): treat all-proxy-models keys as unrestricted in /health A key granted all model permissions stores the literal "all-proxy-models" marker in its models list. The /health access filter compared that marker against real model_names, so the model list filtered down to nothing and the WebUI health check returned healthy_count=0, unhealthy_count=0 with HTTP 503. Skip the filter (both the live path and the background-cache model_id scoping) when the marker is present, matching how auth_checks treats SpecialModelNames.all_proxy_models. Fixes BerriAI#29744. * fix(health): resolve all-team-models sentinel to the team allowlist Same failure shape as the all-proxy-models case: a key carrying the literal "all-team-models" entry matches no real model_name, so the /health access filter would zero out the model list. Resolve the sentinel to the key's team models when team_id is set, matching get_key_models in model_checks.py. Without a team_id the sentinel stays unresolved and matches nothing, denying rather than widening access, mirroring _resolve_key_models_for_auth_check. * feat(proxy): auto-enable drop_params for Claude Code requests (BerriAI#30218) * feat(proxy): auto-enable drop_params for Claude Code requests Claude Code identifies itself with a claude-cli/<version> user agent and sends Anthropic-specific params (top_k, thinking, etc.) on every request. When the proxy routes those requests to a non-Anthropic provider, the unsupported params fail the call unless drop_params is configured. Detect the Claude Code user agent in add_litellm_data_to_request and default drop_params to true for those requests, without overriding an explicit drop_params value sent by the caller. * feat(proxy): respect operator litellm_settings drop_params over Claude Code default An explicit drop_params in the operator's litellm_settings (true or false) now suppresses the Claude Code user agent default, so an operator who deliberately configured drop_params: false keeps strict param validation for Claude Code clients too. The auto-default only fills the gap when neither the request body nor the config sets a value. * fix(snowflake): migrate to native endpoints with auto-routing for Claude models (BerriAI#29964) * fix(snowflake): migrate to native Cortex REST API endpoints Replaces the legacy /api/v2/cortex/inference:complete endpoint with the native OpenAI-compatible /api/v2/cortex/v1/chat/completions endpoint, fixing error 390142 (Incoming request does not contain a valid payload) when using model: snowflake/<model> in LiteLLM proxy. Changes: - litellm/llms/snowflake/chat/transformation.py: route to native /cortex/v1/chat/completions, remove Snowflake-specific tool_spec payload transformation, remove content_list response handling, add stream to supported params - litellm/llms/snowflake/anthropic/transformation.py (new): SnowflakeCortexAnthropicConfig routes Claude models to /cortex/v1/messages with anthropic-version header and Anthropic->OpenAI response transform - tests: 29 unit tests covering URL routing, auth headers, payload format, and response parsing * fix(snowflake): map max_tokens to max_completion_tokens for native endpoint * fix: handle multi-turn tool conversations and OpenAI→Anthropic tool format conversion - _extract_system_and_messages now preserves tool_calls from assistant messages and converts them to Anthropic tool_use content blocks - tool role messages are converted to user role with tool_result content blocks (as required by Anthropic Messages API) - Added _transform_tools_to_anthropic() to convert OpenAI tool format (type/function/parameters) to Anthropic format (name/input_schema) - Added comprehensive tests for multi-turn tool conversations Addresses review feedback on PR BerriAI#29964 * test: add coverage for malformed JSON and non-string tool arguments * fix(tests): update chat transformation tests for native OpenAI-compatible endpoint * style: apply black formatting * fix: resolve mypy type errors in anthropic transformation * fix: correct mypy type: ignore error codes (attr-defined) * fix: use max_tokens instead of max_completion_tokens for Snowflake endpoint compatibility * refactor: merge Anthropic config into unified SnowflakeConfig with auto-routing - Remove separate SnowflakeCortexAnthropicConfig and anthropic/ directory - SnowflakeConfig now auto-routes based on model name: - Claude models → /messages endpoint (Anthropic format) - All others → /chat/completions endpoint (OpenAI format) - No new provider needed (stays as SNOWFLAKE = 'snowflake') - Tool message transformation for Claude: tool_calls → tool_use blocks, tool role → user with tool_result - OpenAI → Anthropic tool format conversion (parameters → input_schema) - Addresses Greptile feedback about unwired SnowflakeCortexAnthropicConfig * fix: use max_completion_tokens for /chat/completions (Snowflake deprecated max_tokens on this endpoint) * fix(tests): update assertions for Claude auto-routing to /messages endpoint * fix(snowflake): add tool_choice conversion and preserve max_completion_tokens in Anthropic path * fix(snowflake): use ChatCompletionMessageToolCall objects and strip model prefix on OpenAI path * fix(snowflake): collect multiple system messages to prevent guardrail override * chore: remove committed .pyc files and add __pycache__ to .gitignore * fix: remove unused Union import * fix: restore original .gitignore (accidentally replaced in earlier commit) * feat(snowflake): add streaming response handler for both Anthropic and OpenAI SSE formats * fix: remove unused AsyncIterator and Iterator imports * fix: add missing total_tokens to ChatCompletionUsageBlock * fix(snowflake): coalesce consecutive tool results into single user message for Anthropic * fix(snowflake): handle message_start event for streaming input_tokens tracking * fix: evict last deleted model in multi-instance deployments (BerriAI#28608) * fix: evict last deleted model in multi-instance deployments _delete_deployment had an early return when db_models was empty, preventing eviction of the last deleted model during reconciliation. - Remove len(db_models)==0 early return from _delete_deployment - Return None (not []) from _get_models_from_db on DB failure so callers can distinguish a transient failure from a genuinely empty DB - Guard _update_llm_router against None to skip updates on DB failure Fixes BerriAI#28443 * test: remove dead MagicMock assignment in type_mismatch test * fix: update test to pass [] not None to _update_llm_router test_ProxyConfig__update_llm_router_bad_proxy_logging_raises was passing None as new_models to get through to the proxy_logging_obj check, but the None guard we added now returns early before reaching that path. Pass [] instead so the test exercises the intended AttributeError case. Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> * chore: regenerate API types to sync schema.d.ts with proxy OpenAPI spec Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> --------- Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> * fix: invalidate Redis spend counter on /key/reset_spend (BerriAI#29694) * fix: set Redis spend counter to reset_to value on /key/reset_spend Previously, the Redis spend counter was always set to 0.0 after a reset, even when reset_to was a non-zero value (partial reset). This caused the budget to be under-enforced for up to 60 seconds until the counter expired and fell through to the DB. Now the counter is set to the actual reset_to value, so partial resets are reflected correctly and budget enforcement is consistent. * test: update reset_key_spend test to match direct cache set The implementation now sets spend_counter_cache directly instead of calling _invalidate_spend_counter. Update the test to verify the in_memory_cache.set_cache call with the correct key, value, and ttl. --------- Co-authored-by: michaelxer <michaelxer@users.noreply.github.com> * fix: add scaleway models pricing (BerriAI#27659) * fix: Add embeddings support for Scaleway provider * fix: resolve merge conflicts * fix(main): clarify backend route handling for Swagger static assets (BerriAI#30196) * fix(main): clarify backend route handling for Swagger static assets * fix(allowlist): add BACKEND_MOUNT_PATHS for Swagger static assets * fix(voyage): route multimodal embeddings to correct endpoint (BerriAI#30193) * fix(voyage): route multimodal embeddings to correct endpoint * test(voyage): cover multimodal embedding edge cases * test(voyage): cover api key fallback * fix(voyage): raise early on missing api key and malformed image url * test(voyage): cover utils routing and helper * fix(voyage): route supported openai params for multimodal models * style: apply black formatting * fix(ui): infer Azure API version from API base (BerriAI#30204) * fix(ui): infer Azure API version from API base * fix(ui): address Azure API version feedback * Update litellm/llms/snowflake/chat/transformation.py Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> * feat(datadog): add team-scoped Datadog callback support (BerriAI#29947) Enable teams to configure their own Datadog credentials via POST /team/{team_id}/callback, following the same pattern as Langfuse. * Merge pull request BerriAI#29528 from aanchal22/litellm_byok-alias-merge fix(proxy): atomic merge for team model aliases and team.models on BYOK create * feat: add EmpirioLabs as an OpenAI-compatible provider (BerriAI#30278) Co-authored-by: Adam Dalloul <adam.d.developer@gmail.com> * fix: resolve failing tests and lint in snowflake/team endpoints - Black-format snowflake/chat/transformation.py to fix lint failure - Update Anthropic config test to expect default max_tokens of 4096 (matches implementation) - Add AsyncMock + execute_raw mock to team_model_add cache-refresh pin test - Add model_dump mock and patch cache/logging in test_uses_atomic_array_append_with_dedup * fix(test): update test_db_error_new_model_check for new _delete_deployment logic _delete_deployment no longer short-circuits on empty db_models — it now treats [] as a valid empty-DB state and proceeds to check config models. Mock get_config to return the two router deployments so they appear in combined_id_list and are protected, which matches the real-world scenario where a DB error occurs but the models are config-backed. * feat(proxy): register cancel_on_disconnect in ConfigGeneralSettings and config list (BerriAI#30295) * feat(proxy): register cancel_on_disconnect in ConfigGeneralSettings and config list Follow-up to BerriAI#30223 per maintainer review: documents the flag in ConfigGeneralSettings with a short description and adds it to allowed_args in get_config_list so the UI and /config/list expose it. A test pins that /config/list returns the field with type Boolean, which requires both registrations to be present * chore(ui): regenerate schema.d.ts for cancel_on_disconnect --------- Co-authored-by: kursad <kursad.lacin@brado.net> * fix(datadog): never fall back to env DD_API_KEY for caller-supplied destinations Team/key-scoped Datadog loggers could be pointed at an arbitrary dd_agent_host or dd_site while omitting dd_api_key, causing the proxy's global DD_API_KEY to be sent as the DD-API-KEY header to that destination. Gate the env-var fallback behind an allow_env_credentials flag, set to False when the destination is caller-supplied, mirroring the existing langfuse/langsmith pattern. --------- Signed-off-by: Rudra Dudhat <contact.rdudhat@gmail.com> Co-authored-by: Emerson Gomes <emerson.gomes@thalesgroup.com> Co-authored-by: daitran-tensormesh <dai@tensormesh.ai> Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com> Co-authored-by: Muspi Merol <me@promplate.dev> Co-authored-by: fangkang <fangkangm@gmail.com> Co-authored-by: Chris Hoogeboom <chris.hoogeboom@gmail.com> Co-authored-by: kursadlacin <kursadlacin@gmail.com> Co-authored-by: kursad <kursad.lacin@brado.net> Co-authored-by: CreateRandom <18438707+CreateRandom@users.noreply.github.com> Co-authored-by: hcl <chenglunhu@gmail.com> Co-authored-by: Filippo Menghi <113345637+Cyberfilo@users.noreply.github.com> Co-authored-by: Mateo Wang <277851410+mateo-berri@users.noreply.github.com> Co-authored-by: sfc-gh-nashukla <navnit.shukla@snowflake.com> Co-authored-by: Rudra Dudhat <contact.rdudhat@gmail.com> Co-authored-by: Michael <52305679+michaelxer@users.noreply.github.com> Co-authored-by: michaelxer <michaelxer@users.noreply.github.com> Co-authored-by: Quentin Champenois <26109239+Quentinchampenois@users.noreply.github.com> Co-authored-by: mauriceberentsen <mauriceberentsen@live.nl> Co-authored-by: lost9999 <56498264+lost9999@users.noreply.github.com> Co-authored-by: GaetanVDB07 <86427581+GaetanVDB07@users.noreply.github.com> Co-authored-by: Aanchal Khandelwal <aan2210khandelwal@gmail.com> Co-authored-by: Adam Dalloul <adam_dalloul@icloud.com> Co-authored-by: Adam Dalloul <adam.d.developer@gmail.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Relevant issues
Fixes multi-turn tool conversation failures on Snowflake Cortex REST API endpoints (customer-reported:
"unknown role tool"on v1.88.0,"Each toolUse block must be accompanied with a matching toolResult block"on v1.82.6).Pre-Submission checklist
make test-unit@greptileaiand received a Confidence Score of at least 4/5 before requesting a maintainer reviewScreenshots / Proof of Fix
Bug reproduction: v1.88.0 sends
role: "tool"directly to Snowflake → rejectedFix: PR transforms tool messages for Claude, passes through for others
Live testing: 12/14 Snowflake Cortex models passing
Streaming verified (both paths)
Token limit handling
Snowflake's
/chat/completionsendpoint rejectsmax_tokensand requiresmax_completion_tokens:Both
max_tokensandmax_completion_tokenslitellm params are accepted from callers and sent asmax_completion_tokensto Snowflake's/chat/completions. The Anthropic/messagespath usesmax_tokens(as Anthropic requires).Design Decisions
max_completion_tokenson OpenAI pathmax_tokens— confirmed via direct API test"choices"vs"type"key — works for both endpointsType
Bug fix + Feature (native endpoint migration)