feat: litellm oss 110626 - #30202

Merged

mateo-berri merged 37 commits into

litellm_internal_stagingfrom

litellm_oss_110626

Jun 12, 2026

Commits on Jun 11, 2026

Add gpt-realtime-whisper Realtime transcription support (OpenAI + Azure) (#29775)

emerzon
authored and
Sameerlite
committed
d042433
Lazy evaluate debug log in pass-through endpoint (#30177)

nina-hu
authored and
Sameerlite
committed
1b54a77
fix(parallel_ai): migrate search integration from v1beta to v1 endpoint (#30157)

authored and
Sameerlite
committed
74bff7e
feat(focus): add Mavvrik destination for FOCUS export (#29935)

pghuge-cloudwiz
authored and
Sameerlite
committed
516a4e7
fix: preserve responses streaming flag (#30189)
alex107ivanov
authored
0dee014
fix(spend/daily-activity): stable offset pagination via id tiebreaker (#30164) (#30167)
hclsys
authored
8929445
fix(oci): inject a default maxTokens so omitted max_tokens doesn't truncate responses (#30018)
fede-kamel
authored
3ff2742
fix(oci): translate response_format json_schema to OCI's accepted shape (#29691)

fede-kamel
and
Sameerlite
authored
460f45c
fix(oci): accept default n=1 on Cohere instead of hard-failing (#29705)

fede-kamel
and
Sameerlite
authored
41924f6
fix(oci): drop max_retries instead of hard-failing on OCI (#29727)

fede-kamel
and
Sameerlite
authored
e1f3a96
fix(spend-logs): rehydrate metadata JSONB text on ui_view_spend_logs (#29682)
hclsys
authored
adf7c97
fix(proxy): release max_parallel_requests slot when a stream is cancelled mid-flight (#27955) (#30020)
Ar-maan05
authored
35dc441
fix(mcp): propagate model into model_call_details for passthrough tool calls (#30122)
auggie246
authored
abf04d0
fix(mcp): surface upstream challenges for delegated OAuth (#30124)
jstar0
authored
b9c0a35
perf(benchmarks): add CPU timing metrics to streaming benchmark (#29980)
SakshamMaggo
authored
aad303d
fix(gemini): don't emit empty choices on metadata-only stream chunks (#29167)
hclsys
authored
e89bd5c
fix(key): allow /key/update to clear budget_limits with [] or null (#30085)
Cyberfilo
authored
93688ea
fix(proxy): persist guardrail info in spend logs for /v1/responses (#30092)
Cyberfilo
authored
d134748
fix(proxy): handle non-standard SSE frames in Anthropic passthrough logging (#26000)

authored
8dfbe86
feat(newrelic): Add New Relic extension (#26989)

bonczj
and
claude
authored
45f9e26
Add Azure AI Foundry DeepSeek V3.1 and V4 Pro/Flash global pricing to cost map (#30134)

mdturp
and
cursoragent
authored
c5df3f2
fix(logging): translate Responses bridge result to ModelResponse for spend logs (#28985)
piotrminkina
authored
706ea8f
fix(anthropic-adapter): re-emit first delta on streaming content-block transitions (#30024)

authored
4782aa9
feat: add opt-in healthy_only filter to GET /v1/models (#30130)

authored
9dd9d23
Dedupe team soft budget alerts by team_id instead of token (#30097)
Cyberfilo
authored
7f17115
feat(bedrock guardrails): support contextual grounding qualifiers (request-side) (#30057)
jpv-costa
authored
c66a4d0
Feat(guardrail): Adding support for custom Ovalix guardrail (#21887)
shalom-ovalix
authored
dcfd2e9
fix(github_copilot): normalize per-event item_id in /responses streaming (#30072)
codgician
authored
33985f7
feat: add /model/block and /model/unblock endpoints (#30125)

AgentGymLeader
and
Sameerlite
authored
fad4689
fix: add week unit support to get_next_standardized_reset_time (#30100)
AgentGymLeader
authored
ca556e3
fix: black formatting and remove undocumented MAVVRIK_FOCUS_FREQUENCY env var
Sameerlite
committed
e267e39
fix: black formatting with correct version and sync schema.d.ts for healthy_only param
Sameerlite
committed
7444774
fix: resolve mypy errors and add transcription_sessions to JSON schema endpoint enum
Sameerlite
committed
5df17c3
fix: restore MAVVRIK_FOCUS_FREQUENCY guard and exclude it from docs key scan
Sameerlite
committed
a30a0a4
fix: address Greptile P2 comments - move constant, use UTC datetime, skip redundant team lookup
Sameerlite
committed
e0684c2
revert: restore original team lookup logic in can_key_call_resolved_model
Sameerlite
committed
3126c22

Commits on Jun 12, 2026

Merge branch 'litellm_internal_staging' into litellm_oss_110626
mateo-berri
committed
5e61498