BerriAI/litellm v1.80.12-nightly

on GitHub

What's Changed

• Add image tokens in chat completion by @Sameerlite in #18327
• Add usage object in image generation of gemini by @Sameerlite in #18328
• [Feat] New provider TTS - Add AWS polly API for TTS by @ishaan-jaff in #18326
• Litellm content filter logs page by @ishaan-jaff in #18335
• fix: request body for image embedding request by @Sameerlite in #18336
• [Fix] Minor Styling Changes by @yuneng-jiang in #18310
• [Feature] UI - Improve Create and Delete Path for CloudZero by @yuneng-jiang in #18263
• [Feat] Enable async_post_call_failure_hook to transform error responses by @AlexsanderHamir in #18348
• [Feature] SSO Role Mapping by @yuneng-jiang in #18090
• [Fix] Allow Deleting Key Expiry by @yuneng-jiang in #18278
• [Docs] Add CloudZero UI Docs by @yuneng-jiang in #18350
• [Infra] Migrate MCP Fetching Logic to React Query by @yuneng-jiang in #18352
• Fix CI: Revert security scan changes and add GitGuardian ignore rules by @AlexsanderHamir in #18358
• [Feature] Allow Error code filtering on Spend Logs by @yuneng-jiang in #18359
• feat(databricks): Add enhanced authentication, security features, and custom user-agent support by @prasadkona in #18349
• [Feature] UI - Add Error Code Filtering on UI by @yuneng-jiang in #18366
• [Fix] UI - Key Creation MCP Settings Submit Form Unintentionally by @yuneng-jiang in #18355
• fix(gemini): properly catch context window exceeded errors by @lucasrothman in #18283
• fix: lost tool_calls when streaming has both text and tool_calls by @Harshit28j in #18316
• fix: remove deprecated Groq models and update model registry by @Chesars in #18062
• Add 5 AI providers using openai_like by @mcowger in #18362
• feat(pricing): add Azure gpt-image-1.5 pricing to cost map by @Chesars in #18347
• docs(openai): fix gpt-5-mini reasoning_effort supported values by @Chesars in #18346
• Add azure_ai/gpt-oss-120b model pricing details by @aggeentik in #18317
• Add supports_response_schema to all supported together ai models by @Sameerlite in #18368
• (Feat) Minimax chat completion support by @Sameerlite in #18380
• Add anthropic native endpoint support for Minimax by @Sameerlite in #18377
• Add gemini thought signature support via tool call id by @Sameerlite in #18374
• Add support for minimax tts by @Sameerlite in #18334
• add sap creds for list in proxy ui by @vasilisazayka in #18375
• [Feat] Interactions API - allow using all litellm providers (interactions -> responses api bridge) by @ishaan-jaff in #18373
• [Feat] RAG query endpoint - Add RAG Search / Query endpoint by @ishaan-jaff in #18376
• [Refactor] Consolidate lazy import handlers with registry pattern by @AlexsanderHamir in #18389
• [Refactor]: Complete lazy loading migration for all 180+ LLM config classes by @AlexsanderHamir in #18392
• [Fix] UI - Disappears in Development Environments by @yuneng-jiang in #18393
• feat: support MCP stdio header env overrides by @uc4w6c in #18324
• [Refactor]: Lazy load additional components (types, callbacks, utilities) by @AlexsanderHamir in #18396
• Revert "[Fix] UI - Disappears in Development Environments" by @yuneng-jiang in #18398
• [Fix] Disable Admin UI Flag by @yuneng-jiang in #18397
• [Fix] Allow Organization Admins to See Organization Tab by @yuneng-jiang in #18400
• [Reissue: Fix] UI - Disappears in Development Environments by @yuneng-jiang in #18399
• [Feature] Resolve Organization Alias on Team Table by @yuneng-jiang in #18401
• (fix) Remove double imports in main.py by @Sameerlite in #18406
• Add generate content in llm route by @Sameerlite in #18405
• feat(vertex-ai): add centralized get_vertex_base_url() helper for global location support by @Sameerlite in #18410
• Add support for image generation via azure ad token by @Sameerlite in #18413
• Update minimax docs to be in proper format by @Sameerlite in #18403
• [Feature] Resolve Team Alias in Organization Info View + Deprecate useTeams by @yuneng-jiang in #18404
• [Refactor] useQuery Hooks Now Depend on useAuthorized by @yuneng-jiang in #18416
• [Fix] Allow Organization Admins to View Their Organization Info by @yuneng-jiang in #18417
• [Fix] UI - Model Page Sorting Sorts Entire Set by @yuneng-jiang in #18420
• [Feature] Allow Column Resizing on Keys Table by @yuneng-jiang in #18424
• [Infra] Add Model Page Model Provider Select Performance by @yuneng-jiang in #18425
• [Feat] AI Gateway - Add support for Platform Fee / Margins by @ishaan-jaff in #18427
• Add docs as follow on to #18362 by @mcowger in #18388
• Add azure/gpt-5.2-chat by @mihidumh in #18361
• add license endpoint by @bernata in #18311
• fix: Datadog span kind fallback when parent_id missing by @uc4w6c in #18418
• feat: add MCP test support to completions on Playground by @uc4w6c in #18440
• added google image gen models by @mubashir1osmani in #18472
• Litellm feat guardrail log actual event type by @uc4w6c in #18489
• fix health status by @uc4w6c in #18443
• Fix formatting in proxy configs documentation by @Prazeina in #18498
• Require auth for MCP connection test endpoint by @emerzon in #18290
• [Feature] Add Optional Query Param "expand" to /key/list by @yuneng-jiang in #18502
• [Refactor] Key Table to Use Optional "expand" query by @yuneng-jiang in #18503
• [Refactor] UI - Keys Table by @yuneng-jiang in #18504
• [Infra] UI - Virtual Keys Table to use React Query by @yuneng-jiang in #18506
• [Infra] UI - E2E Tests Setup by @yuneng-jiang in #18501
• [Feature] UI - Key Table Loading Skeleton by @yuneng-jiang in #18527
• [Refactor] UI - E2E Tests: Sidebar Navigation by @yuneng-jiang in #18528
• feat: optimize MCP server listing by separating health checks by @uc4w6c in #18530
• [Feature] UI - Add Model Clearer Flow For Team Admins by @yuneng-jiang in #18532
• [Refactor] UI - E2E Tests: Parity With Legacy Tests by @yuneng-jiang in #18533
• [Feature] UI - New Badge Reusable Component + Update New badge for features by @yuneng-jiang in #18537
• [Feauture] UI - Hide New Badges by @yuneng-jiang in #18547
• [Infra] UI - Unit Testing Coverage With v8 by @yuneng-jiang in #18548
• [Refactor] UI - Remove Model Analytics From Model Page by @yuneng-jiang in #18552
• [Fix] UI - Usage Page User Max Budget by @yuneng-jiang in #18555
• fix: prevent LiteLLM from closing external OTEL spans by @uc4w6c in #18553
• [Feature] Expire Previous UI Session Tokens on Login by @yuneng-jiang in #18557
• [Fix] CI/CD - check_code_and_doc_quality by @AlexsanderHamir in #18560
• [Infra] CI - De-duplicate Docker Database Build by @yuneng-jiang in #18561
• [Fix] CI/CD - litellm_mapped_tests_llms by @AlexsanderHamir in #18563
• [Fix] CI/CD - litellm_utils_testing by @AlexsanderHamir in #18565
• [Fix] CI/CD - litellm_security_tests by @AlexsanderHamir in #18567
• [Fix] CI/CD - litellm_mapped_tests_proxy by @AlexsanderHamir in #18572
• [Feature] UI - Model Page Loading State by @yuneng-jiang in #18574
• [Feature] UI - Clicking on Logo Directs to Correct URL by @yuneng-jiang in #18575
• [Refactor] UI - Model Hub Page by @yuneng-jiang in #18568
• [Fix] UI - Change Budget page to Have Tabs by @yuneng-jiang in #18576
• [Refactor] UI - Test for React Query Hooks by @yuneng-jiang in #18577
• fix authentication errors at messages API via azure_ai by @systec-dk in #18500
• Add Kubernetes ServiceAccount JWT authentication support by @LukasdeBoer in #18055
• fix: unify model names to provider-defined names by @uc4w6c in #18573
• add selectable mcp servers to the playground by @uc4w6c in #18578
• feat: add UI support for configuring meta URLs by @uc4w6c in #18580
• fix(vertex_ai): convert image URLs to base64 for Vertex AI Anthropic by @DevajMody in #18497
• fix(vertex_ai): separate Tool objects for each tool type per API spec by @xuan07t2 in #18514
• feat: Add MiniMax provider support to UI dashboard by @Chesars in #18496
• Fix Gemini 3 imgs in tool response by @RensDimmendaal in #18190
• fix: correct deepseek-v3p2 pricing for Fireworks AI by @yurekami in #18483
• Add all resolution for gpt-image-1.5 by @Sameerlite in #18586
• Preserve system instructions for gemini by @Sameerlite in #18585
• Add thought signature for non tool call requests by @Sameerlite in #18581
• Remove prompt caching headers as the support has been removed by @Sameerlite in #18579
• Add validation for negative budget by @Sameerlite in #18583
• fix background cost tracking tests by @Sameerlite in #18588
• fix(cost_calculator): correct gpt-image-1 cost calculation using token-based pricing by @Chesars in #17906
• Feature: Add log_format option to GenericAPILogger by @chetanchoudhary-sumo in #18587
• fix: return empty data array instead of 500 when no models configured by @agertz7 in #18556
• fix: add deprecation_date for discontinued Google models by @yudelevi in #18550
• fix: logging_worker handle event loop changes in multiprocessing by … by @Harshit28j in #18423
• Add memory pattern detection test and fix bad memory patterns by @AlexsanderHamir in #18589
• add dimensions for ollama embedding by @smallp in #18536
• Add unbounded data structure detection to memory test by @AlexsanderHamir in #18590
• refactor: Add lazy loading for get_llm_provider by @AlexsanderHamir in #18591
• Feat/add watsonx fields by @kevinpauer in #18569
• refactor: lazy-load heavy audio library and loggers by @AlexsanderHamir in #18592
• Lazy load 9 heavy imports in litellm/utils.py by @AlexsanderHamir in #18595
• fix openapi to mcp by @uc4w6c in #18597
• [Feature] UI - Allow Adding Role Mappings for SSO by @yuneng-jiang in #18593
• [Feature] UI - SSO Settings Page by @yuneng-jiang in #18600
• [Infra] UI - Adding Unit Tests to Increase Coverage by @yuneng-jiang in #18601
• [Fix] UI - Useful Links Remove Modal on Adding Links by @yuneng-jiang in #18602
• Agentcore mcp server docs by @krrishdholakia in #18603
• LiteLLM adopters section by @krrishdholakia in #18605
• fix: correct cost calculation when reasoning_tokens are without text_tokens by @Chesars in #18607
• Iam roles anywhere docs by @krrishdholakia in #18559
• Litellm feat helm lifecycle support by @cansakiroglu in #18517
• fix(sap): pass through extra params from allowed_openai_params by @dee-walia20 in #18432
• fix: handle empty error objects in response conversion by @yurekami in #18493
• feat(types): add output_text property to ResponsesAPIResponse by @yurekami in #18491
• google genai adapter inline data support by @luxinfeng in #18477
• feat(zai): Add GLM-4.7 model with reasoning support by @cantalupo555 in #18476
• Allow installation with current grpcio on old Python by @andersk in #18473
• fix: correctly route codestral chat and FIM endpoints by @majiayu000 in #18467
• fix: extract pure base64 data from data URLs for Ollama by @majiayu000 in #18465
• fix: align prometheus metric names with DEFINED_PROMETHEUS_METRICS by @majiayu000 in #18463
• refactor: lazy load get_llm_provider and remove_index_from_tool_calls by @AlexsanderHamir in #18608
• [Refactor] UI - SSO Settings Page Edit Path by @yuneng-jiang in #18609
• refactor(utils): lazy load heavy imports to improve import time and memory usage by @AlexsanderHamir in #18610
• refactor(utils): implement lazy loading for provider configs, model info classes, streaming handlers, and redact utilities by @AlexsanderHamir in #18611
• [Infra] UI - E2E Tests: Add Model - See Models from Selected Provider by @yuneng-jiang in #18615
• refactor(utils): lazy load 15 additional imports to improve import time by @AlexsanderHamir in #18613
• perf(utils): lazy load 15+ unused imports by @AlexsanderHamir in #18616
• [Feature] UI - SSO Settings Loading State + Deprecate Previous SSO Flow by @yuneng-jiang in #18617
• [Infra] UI - Unit Tests: Adding Tests to Increase Coverage by @yuneng-jiang in #18618
• [Feature] UI - Virtual Keys Table Loading State Between Pages by @yuneng-jiang in #18619
• feat: Add Levo AI integration by @amangupta-20 in #18529
• fix: MCP handling in unified guardrail by @uc4w6c in #18630
• refactor: remove redundant Bedrock guardrail block handling by @uc4w6c in #18634
• feat: add user email to cloudzero by @uc4w6c in #18584
• Fix: Remove exec() usage and handle invalid OpenAPI parameter names by @hamzaq453 in #18480
• [feat] mcp global mode by @uc4w6c in #18639
• tests: mock LiteLLM_UserTable for CloudZero export by @uc4w6c in #18640
• Add openrouter/google/gemini-3-flash-preview to model_prices_and_context_window.json by @matt-greathouse in #18247
• [Feat] Add Cost Estimator for AI Gateway by @ishaan-jaff in #18643
• feat(mcp): parallelize tool fetching from multiple MCP servers by @ktsaou in #18627
• fix: Add custom llm provider to get_llm_provider when sent via UI by @Sameerlite in #18638
• fix: auth header for custom api base in generate Content request by @Sameerlite in #18637
• (feat) Add mapping for reasoning effort to summary param of responses API by @Sameerlite in #18635
• Fix mapped tests 05012026 by @Sameerlite in #18644
• [UI] Add view for estimating costs across requests by @ishaan-jaff in #18645
• [Feat] Litellm UI allow selecting many models for cost estimator by @ishaan-jaff in #18653
• lasso-guardrail-api docs by @orgersh92 in #18652
• refactor: lazy load DatadogLLMObsInitParams to reduce import time by @AlexsanderHamir in #18658
• refactor: migrate utils.py lazy imports to registry pattern by @AlexsanderHamir in #18657
• fix: prevent proxy startup failures from case-sensitive tool permission guardrail validation by @AlexsanderHamir in #18662
• Generic guardrail api update by @krrishdholakia in #18647
• [Feat] Add support for LangSmith organization-scoped API keys with tenant ID by @shivamrawat1 in #18623
• fix(router): Validate routing_strategy at startup to fail fast with helpful error by @FlibbertyGibbitz in #18624
• Add libsndfile to database Docker image for audio processing by @Chesars in #18612
• fix: extend case normalization to ALL guardrail types by @AlexsanderHamir in #18664
• Fix: Map Gemini cached_tokens to Langfuse cache_read_input_tokens by @Chesars in #18614
• feature/ add qualifire guardrails by @drorIvry in #18594
• Fix User Invite & Key Generation Email Notification Logic by @urainshah in #18524
• [Fix] Use timedelta to Calculate Key Expiry on Generate by @yuneng-jiang in #18666
• [docs] mention MCP prompt/resources types in overview by @uc4w6c in #18669
• [feat] add support MCP guardrail to noma by @uc4w6c in #18668
• fix: ensure_async_clients_test error by @uc4w6c in #18670
• [Feature] UI - SSO Settings Page Add Role Mappings by @yuneng-jiang in #18677
• Use already configured opentelemetry providers by @mangabits in #18279
• [feat] mcp server visibility configurable by @uc4w6c in #18681
• [Fix] UI - SSO Edit Modal Clear Role Mapping Values on Provider Change by @yuneng-jiang in #18680
• Fix: fixed some MCP gateway oauth2 auth issues and ClosedResourceError by @Eric84626 in #18281
• feat: Add GigaChat provider support by @0717376 in #18564
• Add redisvl in requirements.txt by @Sameerlite in #18177
• Staging 01/05/2026 by @krrishdholakia in #18663
• feat (proxy) : Responses API Support More Streaming Callback Hooks by @LingXuanYin in #18513
• feat(model_cost): add display_name, model_vendor, and model_version metadata to model entries by @nmgarza5 in #17330
• Revert "Add redisvl in requirements.txt" by @uc4w6c in #18687
• fix main test by @uc4w6c in #18693
• Fix mapped test 060126 by @Sameerlite in #18695
• Revert "feat(model_cost): add display_name, model_vendor, and model_version metadata to model entries" by @Sameerlite in #18701
• [Feat] New API Endpoint - Responses API (v1/responses/compact) by @ishaan-jaff in #18697
• fix: remove display name by @Sameerlite in #18702
• [Fix] Put automatic reasoning summary inclusion behind feat flag by @Sameerlite in #18688
• Add the LITELLM_REASONING_AUTO_SUMMARY in doc by @Sameerlite in #18705
• [Bug] fix model map by @Sameerlite in #18706
• [UI] - Feat add request provider form on UI by @ishaan-jaff in #18704
• fix: preserve client error status codes in streaming mode #18690 by @krisxia0506 in #18698
• Fix/gcs cache docs missing for proxy mode by @pascalwhoop in #13328
• [Feature]: Add amazon.nova-2-multimodal-embeddings-v1 to model_prices_and_context_window by @Isaac4real in #18710
• Remove redundant comments about setting litellm.callbacks by @marukaz in #18711
• fix(router): correct num_retries tracking in retry logic by @rohitravirane in #18712
• Add client header for SAP AI Core Tracking by @lizzzcai in #18714
• Add LlamaGate as a new provider by @hkd987 in #18673
• Handle not supported region for vertex ai count tokens - v1/messages/count_tokens by @otaviofbrito in #18665
• Add LITELLM_DISABLE_LAZY_LOADING env var to fix VCR cassette creation issue by @AlexsanderHamir in #18725
• [Docs]: Adding SigNoz integration to observability docs by @gkarthi-signoz in #18726
• [Infra] Fixing UI Build by @yuneng-jiang in #18732
• [Feature] Allow editing team_member_budget_duration in /team/update by @yuneng-jiang in #18735
• [Feature] UI - Reusable Duration Select + Team Update Member Budget Duration by @yuneng-jiang in #18736
• [feat] allow configuring arize project name for OpenTelemetry service name by @uc4w6c in #18738
• [Infra] UI - Adding Unit Test for Coverage by @yuneng-jiang in #18742
• [Infra] UI - E2E Test: Can View Admin Setting Page by @yuneng-jiang in #18682
• Litellm fixes a2a sdk missing in requirements.txt by @ishaan-jaff in #18748
• [Fix] A2a endpoint - fix timeout by @ishaan-jaff in #18747
• [Feat] Litellm new endpoint add container file upload by @ishaan-jaff in #18743
• Revert "Litellm fixes a2a sdk (#18748)" by @ishaan-jaff in #18752
• fix: test_azure_image_edit_cost_tracking by @Sameerlite in #18758
• Litellm improve endpoint discovery by @krrishdholakia in #18762
• Fix component label automation to prevent false positives by @Chesars in #18765

New Contributors

• @prasadkona made their first contribution in #18349
• @lucasrothman made their first contribution in #18283
• @aggeentik made their first contribution in #18317
• @mihidumh made their first contribution in #18361
• @Prazeina made their first contribution in #18498
• @systec-dk made their first contribution in #18500
• @xuan07t2 made their first contribution in #18514
• @RensDimmendaal made their first contribution in #18190
• @yurekami made their first contribution in #18483
• @agertz7 made their first contribution in #18556
• @yudelevi made their first contribution in #18550
• @smallp made their first contribution in #18536
• @kevinpauer made their first contribution in #18569
• @cansakiroglu made their first contribution in #18517
• @dee-walia20 made their first contribution in #18432
• @luxinfeng made their first contribution in #18477
• @cantalupo555 made their first contribution in #18476
• @andersk made their first contribution in #18473
• @majiayu000 made their first contribution in #18467
• @amangupta-20 made their first contribution in #18529
• @ktsaou made their first contribution in #18627
• @FlibbertyGibbitz made their first contribution in #18624
• @urainshah made their first contribution in #18524
• @mangabits made their first contribution in #18279
• @0717376 made their first contribution in #18564
• @nmgarza5 made their first contribution in #17330
• @Isaac4real made their first contribution in #18710
• @marukaz made their first contribution in #18711
• @rohitravirane made their first contribution in #18712
• @lizzzcai made their first contribution in #18714
• @hkd987 made their first contribution in #18673
• @gkarthi-signoz made their first contribution in #18726

Full Changelog: v1.80.11-nightly...v1.80.12-nightly