[Feat] Cancel upstream LLM request on client disconnect by CreateRandom · Pull Request #25776 · BerriAI/litellm

CreateRandom · 2026-04-15T12:42:15Z

Relevant issues

ℹ️ This is a re-implementation of #14295, which was merged then reverted due to a CI failure
caused by an unrelated uvicorn version bump in that PR. This version does not touch any
dependencies.

Pre-Submission checklist

Please complete all items before asking a LiteLLM maintainer to review your PR

I have Added testing in the tests/test_litellm/ directory, Adding at least 1 test is a hard requirement - see details
My PR passes all unit tests on make test-unit
My PR's scope is as isolated as possible, it only solves 1 specific problem
I have requested a Greptile review by commenting @greptileai and received a Confidence Score of at least 4/5 before requesting a maintainer review

Delays in PR merge?

If you're seeing a delay in your PR being merged, ping the LiteLLM Team on Slack (#pr-review).

CI (LiteLLM team)

CI status guideline:

50-55 passing tests: main is stable with minor issues.

45-49 passing tests: acceptable but needs attention

<= 40 passing tests: unstable; be careful with your merges and assess the risk.

Branch creation CI run
Link:
CI run for the last commit
Link:
Merge / cherry-pick CI run
Links:

Screenshots / Proof of Fix

Non-Streaming use case

non-streaming.mp4

Bottom-left: vLLM request is canceled in flight once cancelled by client via litellm

Streaming use case

streaming.mp4

Bottom-left: vLLM request is also canceled in flight once cancelled by client via litellm

Type

🆕 New Feature

Changes

When a client disconnects mid-request, the proxy now cancels the in-flight upstream LLM API call
rather than letting it run to completion.

What changed vs #14295:

No uvicorn version bump (probably the cause of the original CI failure)
Uses request.receive() + "http.disconnect" message type instead of request.is_disconnected(), which is more reliable across uvicorn versions
CancelledError is re-raised naturally instead of being converted to HTTPException(499) — this bypasses the failure hook pipeline (no spurious failure metrics) and lets the ASGI framework handle it silently, which
is the correct behaviour when the client is simply gone
_check_request_disconnection moved from proxy_server.py to common_request_processing.py, next to its only call site
DEFAULT_CLIENT_DISCONNECT_CHECK_TIMEOUT_SECONDS is env-configurable (default 600s)
Test added at tests/test_litellm/proxy/test_client_disconnection.py

vercel · 2026-04-15T12:42:23Z

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
litellm	Ready	Preview, Comment	Apr 16, 2026 7:56am

CreateRandom · 2026-04-15T12:43:49Z

@greptileai

CLAassistant · 2026-04-15T12:43:56Z

All committers have signed the CLA.

greptile-apps · 2026-04-15T12:50:29Z

Greptile Summary

This PR wires up client-disconnect cancellation for non-streaming requests: a _check_request_disconnection task monitors the ASGI receive() channel and cancels the in-flight asyncio.gather future when http.disconnect arrives. A disconnect_event flag lets the CancelledError handler distinguish a true client hangup (→ HTTPException(499), with early-return in _handle_llm_api_exception that skips post_call_failure_hook) from server-shutdown cancellations (→ re-raise). The old dead-code check_request_disconnection in proxy_server.py is removed.

Confidence Score: 5/5

Safe to merge — all prior P1 concerns are addressed and no new blocking issues found.

All four previously-threaded concerns (streaming comment, asyncio.sleep patching in the disconnect test, CancelledError source disambiguation, and post_call_failure_hook still firing for 499) are resolved. The implementation correctly uses request.receive() for disconnect detection, the disconnect_event disambiguates server shutdown from client hangup, _handle_llm_api_exception exits before the failure hook for 499, and the test patches asyncio.sleep. No new P0 or P1 findings were identified.

No files require special attention.

Important Files Changed

Filename	Overview
litellm/proxy/common_request_processing.py	Adds `_check_request_disconnection` and wires it into `async_completion` using `asyncio.gather` cancellation + `disconnect_event`; updates `_handle_llm_api_exception` to exit early (before `post_call_failure_hook`) for HTTP 499; also includes unrelated Black-formatting-only changes.
litellm/proxy/proxy_server.py	Removes the old, never-called `check_request_disconnection` dead-code function; no other logic changes.
litellm/constants.py	Adds `DEFAULT_CLIENT_DISCONNECT_CHECK_TIMEOUT_SECONDS` (default 600 s, env-configurable) for the new disconnect polling loop.
tests/test_litellm/proxy/test_client_disconnection.py	New test file with two async unit tests: one verifying task cancellation on disconnect (patches asyncio.sleep), one verifying no cancellation during normal operation; all mock-based, no network calls.

Sequence Diagram

sequenceDiagram
    participant Client
    participant ASGI as FastAPI/ASGI
    participant ACP as async_completion
    participant DC as _check_request_disconnection
    participant LLM as Upstream LLM

    Client->>ASGI: POST /chat/completions
    ASGI->>ACP: await async_completion(request)
    ACP->>LLM: asyncio.gather llm_call
    ACP->>DC: asyncio.create_task

    par Upstream call succeeds
        LLM-->>ACP: response
        ACP->>DC: disconnect_task.cancel()
        ACP-->>ASGI: return response
        ASGI-->>Client: 200 OK
    and Client disconnects mid-request
        DC->>DC: asyncio.sleep(1)
        DC->>ASGI: await request.receive()
        Client--xASGI: disconnects
        ASGI-->>DC: type http.disconnect
        DC->>ACP: llm_responses.cancel()
        DC->>DC: disconnect_event.set()
        ACP->>ACP: CancelledError caught
        ACP->>LLM: cancel in-flight request
        ACP-->>ASGI: raise HTTPException 499
    end

_{Reviews (4): Last reviewed commit: "style: apply black formatting" | Re-trigger Greptile}

CreateRandom · 2026-04-15T14:18:14Z

@greptileai

CreateRandom · 2026-04-16T07:46:50Z

@greptileai

codecov · 2026-04-16T07:49:39Z

Codecov Report

❌ Patch coverage is 61.53846% with 10 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
litellm/proxy/common_request_processing.py	61.53%	10 Missing ⚠️

📢 Thoughts on this report? Let us know!

CreateRandom · 2026-04-16T08:19:24Z

Lint fails on unrelated files.
Unit tests are also not related to the files that were changed here.

krrish-berri-2 · 2026-05-11T16:31:22Z

i don't think this actually works on streaming - based on my testing i still see the backend running. The current implementation would work for non-streaming, but for streaming it wouldn't as the await is resolved and sent to client

harish-berri · 2026-05-11T16:41:00Z

Hi @CreateRandom — thanks for the PR.

I’ve been thinking about the cancellation approach from the angle of the disconnect handler and how it observes disconnect events. Because it busy-waits on the existing event loop, I’m a bit worried about how that pattern scales when these handlers are created per request and keep polling for a disconnect. I’d also like to understand cleanup better end-to-end—both how the web framework tears down request-scoped resources and how our SDK clears any state and tasks tied to a cancelled request.

That said, I really like the direction and plan to follow up with memory/CPU profiling plus correctness checks on realistic workloads.

Longer term, I’m wondering if something closer to a Go-style request context—propagated through the call stack with timeout/deadline-based cancellation—might give us similar behavior without the polling loop. Totally open to brainstorming that here.

Thanks again for pushing this forward—happy to discuss in more detail whenever works for you.

krrish-berri-2 · 2026-05-11T16:45:15Z

backend streaming still continues: https://cursor.com/artifacts/v/art-cb7f8209-9bc2-4667-9a19-9fc7b23cf934

kursadlacin · 2026-06-11T18:13:51Z

Thanks for this; the receive-based watcher design here is the right call given that BaseHTTPMiddleware breaks is_disconnected polling. I opened #30223 which builds directly on it (credited via Co-authored-by) while addressing the two P1s Greptile raised on the maintainer port in #27146: the watcher is opt-in behind general_settings.cancel_on_disconnect (default off, so the default path is byte-identical and the CI regression that killed #14295 cannot recur), and the 499 path no longer skips post_call_failure_hook, so max_parallel_requests slot release and spend logging keep working on cancellation. Happy for maintainers to take whichever PR is easier to land

…t (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>

* 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>

* 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>

* 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>

feat(proxy): cancel upstream LLM request on client disconnect

26330c8

vercel Bot deployed to Preview April 15, 2026 12:43 View deployment

greptile-apps Bot reviewed Apr 15, 2026

View reviewed changes

Comment thread litellm/proxy/common_request_processing.py Outdated

Comment thread tests/test_litellm/proxy/test_client_disconnection.py

Comment thread litellm/proxy/common_request_processing.py

fix(proxy): address review comments on client disconnect feature

c5f48cc

vercel Bot deployed to Preview April 15, 2026 14:19 View deployment

greptile-apps Bot reviewed Apr 15, 2026

View reviewed changes

Comment thread litellm/proxy/common_request_processing.py

fix(proxy): skip post_call_failure_hook for 499 client disconnect

65bed54

vercel Bot deployed to Preview April 16, 2026 07:47 View deployment

style: apply black formatting

c9c03eb

vercel Bot deployed to Preview April 16, 2026 07:56 View deployment

CreateRandom marked this pull request as ready for review April 16, 2026 08:20

greptile-apps Bot mentioned this pull request May 5, 2026

[WIP] Litellm client disconnect relay. Routing OSS PR into CI #27146

Open

7 tasks

This was referenced May 20, 2026

fix(proxy): close common streaming responses on exit #28353

Closed

fix(proxy): close common streaming responses on exit #28536

Open

kursadlacin mentioned this pull request Jun 11, 2026

fix(proxy): cancel in-flight upstream LLM request on client disconnect (opt-in) #30223

Merged

4 tasks

Sameerlite closed this Jun 12, 2026

Uh oh!

Uh oh!

Conversation

CreateRandom commented Apr 15, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Relevant issues

Pre-Submission checklist

Delays in PR merge?

CI (LiteLLM team)

Screenshots / Proof of Fix

Non-Streaming use case

Streaming use case

Type

Changes

Uh oh!

vercel Bot commented Apr 15, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

CreateRandom commented Apr 15, 2026

Uh oh!

CLAassistant commented Apr 15, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

greptile-apps Bot commented Apr 15, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Greptile Summary

Confidence Score: 5/5

Important Files Changed

Sequence Diagram

Uh oh!

Uh oh!

Uh oh!

Uh oh!

CreateRandom commented Apr 15, 2026

Uh oh!

Uh oh!

CreateRandom commented Apr 16, 2026

Uh oh!

codecov Bot commented Apr 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

CreateRandom commented Apr 16, 2026

Uh oh!

krrish-berri-2 commented May 11, 2026

Uh oh!

harish-berri commented May 11, 2026

Uh oh!

krrish-berri-2 commented May 11, 2026

Uh oh!

kursadlacin commented Jun 11, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

CreateRandom commented Apr 15, 2026 •

edited

Loading

vercel Bot commented Apr 15, 2026 •

edited

Loading

CLAassistant commented Apr 15, 2026 •

edited

Loading

greptile-apps Bot commented Apr 15, 2026 •

edited

Loading

codecov Bot commented Apr 16, 2026 •

edited

Loading