What's Changed
- [Fix] Swap URL Input and Display Name inputs by @yuneng-jiang in #17682
- [Fix] AI Gateway Auth - allow using wildcard patterns for public routes by @ishaan-jaff in #17686
- feat: add Cerebras zai-glm-4.6 model support by @Chesars in #17683
- [Feat] New model - add bedrock writer models by @ishaan-jaff in #17685
- fix(router): handle tools=None in completion requests by @Chesars in #17684
- docs(json): make it clearer how to get Pydantic model output by @yisding in #17671
- mitigate PydanticDeprecatedSince20 warnings by @demoray in #17657
- fix Azure AI Anthropic api-key header and passthrough cost calculation by @emilsvennesson in #17656
- fix(responses): streaming tool_calls being dropped when text + tool_calls by @Chesars in #17652
- fix(openai): include 'user' param for responses API models by @Chesars in #17648
- docs: Add SumoLogic integration documentation by @chetanchoudhary-sumo in #17647
- fix: Use python instead of wget for healthcheck in docker-compose.yml by @kumekay in #17646
- Add 227 new Fireworks AI models by @ishaan-jaff in #17692
- chore: cleanup unused scripts and fix misplaced test file by @Chesars in #17611
- Fix Cerebras context window errors not recognized by @estsauver in #17587
- [Fix] dashscope default base_url error by @chenzhaofei01 in #17584
- fix: resolve UI session MCP permissions across real teams by @uc4w6c in #17620
- [Fix] Change deprecation banner to only show on /sso/key/generate by @yuneng-jiang in #17681
- Update azure search docs by @krrishdholakia in #17726
- Fix error about encoding video id for azure by @Sameerlite in #17708
- Passthrough in response by @dsduenas in #17102
- [Docs] Adding known issues to 1.80.5-stable docs by @yuneng-jiang in #17738
- docs: Add Note for Authentication by @shivamrawat1 in #17733
- [Feature] Get and Update Backend Routes for UI Settings by @yuneng-jiang in #17689
- feat: propagate Langfuse trace_id by @uc4w6c in #17669
- [Fix] Perf - Reduce memory accumulation of spend_logs by @AlexsanderHamir in #17742
- [Fix] Change credential encryption to only affect db credentials by @yuneng-jiang in #17741
- Return 403 instead of 503 for unauthorized routes by @ephrimstanley in #17723
- [Feature] UI - UI Settings by @yuneng-jiang in #17697
- [Fix] Containers API - Allow using LIST, Create Containers using custom-llm-provider by @ishaan-jaff in #17740
- [Feat] Containers API - add new container API file management + UI Interface by @ishaan-jaff in #17745
- feat(voyage): add rerank API support by @Chesars in #17744
- Add community doc link by @krrishdholakia in #17734
- Fix: Extract service_tier from response/usage for OpenAI flex pricing by @shivamrawat1 in #17748
- [docs] add docs for containers files api + code interpreter on LiteLLM by @ishaan-jaff in #17749
- Fix : use litellm params for all videos apis by @Sameerlite in #17732
- Handle gemini context error by @krrishdholakia in #17751
- fix(anthropic): preserve server_tool_use and web_search_tool_result in multi-turn conversations by @Chesars in #17746
- Remove streaming logging doc by @krrishdholakia in #17739
- Support model names with slashes on Gemini endpoints by @hwittenborn in #17743
- fix(docs): Supported Endpoints page by @Chesars in #17710
- fix(guardrails): LiteLLM content filter - mask all regex pattern matches, not just first by @peterkc in #17727
- [Fix] UI SSO - allow upserting user role when SSO provider role changes by @ishaan-jaff in #17754
- Arize Phoenix OSS - Prompt Management Integration by @krrishdholakia in #17750
- fix: Avoid error when we have just the tool_calls in input by @LucasSugi in #17753
- feat: Add nested field removal support to additional_drop_params by @Sameerlite in #17711
- Add support for computer use for gemini by @Sameerlite in #17756
- Add DeepSeek-V3.2 model support by @Sameerlite in #17770
- Fix nvdia and geminin tests by @Sameerlite in #17769
- [Feature] Allow Root Path to Redirect when Docs not on Root Path by @yuneng-jiang in #16843
- add a new feature fix to expose the team alias when authenticating th… by @saisurya237 in #17725
- [Feat] Agent Gateway - Add cost per query for agent invocations by @ishaan-jaff in #17774
- Update model new swagger by @krrishdholakia in #17778
- Document token count endpoint by @krrishdholakia in #17772
- [Feat] Agent Gateway - Add token counting non streaming + streaming by @ishaan-jaff in #17779
- [Feat] New Provider - add langgraph by @ishaan-jaff in #17783
- [Perf] - Cut memory leak in half by @AlexsanderHamir in #17784
- [Feat] A2a gateway - Add cost per token pricing by @ishaan-jaff in #17780
- [Feat] A2a Gateway - allow using bedrock agentcore, langgraph agents with A2a Gateway by @ishaan-jaff in #17786
- fix: suggest Gunicorn instead of uvicorn when using max_requests_before_restart by @AlexsanderHamir in #17788
- [Feat] UI SSO - allow fetching role from generic SSO provider (Keycloak) by @ishaan-jaff in #17787
- [Feat] Agent Gateway - Track
agent_idin SpendLogs by @ishaan-jaff in #17795 - fix: respect videos content db creds by @Sameerlite in #17771
- [Feat] Agent Gateway - allow adding langgraph, bedrock agent core agents by @ishaan-jaff in #17802
- feat(models): add Mistral Codestral 2508, Devstral 2512 models by @Chesars in #17801
- [docs] made litellm proxy and sdk difference cleaner in overview with a table by @shivamrawat1 in #17790
- fix: responses api not applying tpm rate limits on api keys by @raghav-stripe in #17707
- fix: Preserve systemInstructions for vertex ai generate content request by @Sameerlite in #17803
- Add anthropic retrieve batches and retreive file content support by @Sameerlite in #17700
- Fix llm provider for azure_ai in model map by @Sameerlite in #17805
- Add Hiddenlayer Guardrail Hooks by @Ashton-Sidhu in #17728
- fix(azure_ai): Remove unsupported params from Azure AI Anthropic requests by @Chesars in #17822
- feat(bedrock): add serviceTier support for Converse API by @CyrusTC in #17810
- chore: prefer standard trace id for Langfuse logging by @uc4w6c in #17791
- fix(anthropic): capture web_search_tool_result in streaming for multi-turn conversations by @Chesars in #17798
- fix(completion): transform image content in tool results for Responses API by @Chesars in #17799
- fix: MCP OAuth callback routing and URL handling by @uc4w6c in #17789
- [Feature] Sendgrid integration by @yuneng-jiang in #17775
- [Fix] Ensure Model Page Honors UI Settings by @yuneng-jiang in #17804
- [Fix] Create Team Model Dropdown Honors Organization's Models by @yuneng-jiang in #17834
- Add support for OpenAI GPT-5.2 models by @PeterDaveHello in #17836
- [Fix] CI/CD – Clean Up Performance PR Changes & others by @AlexsanderHamir in #17838
- [Fix] CI/CD – Docs & Spend logs by @AlexsanderHamir in #17843
- fix(openai): use optimized async http client for text completions by @Chesars in #17831
- fix: attach team to org table by @mubashir1osmani in #17832
- [Feature] Daily Agent Usage Backend by @yuneng-jiang in #17781
- [Feature] Agent Usage UI by @yuneng-jiang in #17797
- [Fix] Remove SSO Config Values from Config Table on SSO Update by @yuneng-jiang in #17668
- feat: add support for configurable confidence score thresholds and scope in Presidio PII masking by @dominicfallows in #17817
- [Fix] Deduplicate /tag/daily/activity metadata by @yuneng-jiang in #16764
- [Bug Fix] Watsonx Audio Transcription - ensure only correct params are sent to API by @ishaan-jaff in #17840
- [Feat] New API Provider - Add Azure AI Foundry Agents on /chat/completions, /responses, /messages + Agent Gateway by @ishaan-jaff in #17845
- feat(guardrails): add configurable fail-open, timeout, and app_user tracking to panw_prisma_airs guardrail by @jroberts2600 in #17785
- fix: add Python 3.14 support via grpcio version constraints by @peterkc in #17666
- feat(deepseek): add native support for thinking and reasoning_effort params by @Chesars in #17712
- fix: use Union syntax for Python 3.9 compatibility by @Chesars in #17714
- fix: update pricing for global.anthropic.claude-haiku-4-5-20251001-v1:0 by @jichmi in #17703
- [Fix] Missing content in Anthropic to OpenAI conversion by @ultmaster in #17693
- feat(langfuse): Add support for custom masking function by @Chesars in #17826
- Prompt Management - new API for integrating providers by @krrishdholakia in #17829
- [Refactor] UI - Usage Page Components by @yuneng-jiang in #17848
- [Feature] UI - Usage Page View Select by @yuneng-jiang in #17854
- updated docs links for mistral and anthropic by @ryan-crabbe in #17852
- bump openai package to 2.9.0 by @Sameerlite in #17818
- [Feature] Model Hub Useful Links Rearrange by @yuneng-jiang in #17859
New Contributors
- @yisding made their first contribution in #17671
- @emilsvennesson made their first contribution in #17656
- @kumekay made their first contribution in #17646
- @chenzhaofei01 made their first contribution in #17584
- @shivamrawat1 made their first contribution in #17733
- @ephrimstanley made their first contribution in #17723
- @hwittenborn made their first contribution in #17743
- @peterkc made their first contribution in #17727
- @saisurya237 made their first contribution in #17725
- @Ashton-Sidhu made their first contribution in #17728
- @CyrusTC made their first contribution in #17810
- @jichmi made their first contribution in #17703
- @ryan-crabbe made their first contribution in #17852
Full Changelog: v1.80.9.dev1...v1.80.9.dev6