BerriAI/litellm v1.62.1-nightly

on GitHub

What's Changed

• Allow team/org filters to be searchable on the Create Key Page + Show team alias on Keys Table by @krrishdholakia in #8881
• Add created_by and updated_by fields to Keys table by @krrishdholakia in #8885
• (Proxy improvement) - Raise BadRequestError when unknown model passed in request by @ishaan-jaff in #8886
• (Improvements) use /openai/ pass through with OpenAI Ruby for Assistants API by @ishaan-jaff in #8884
• Update model path and documentation for Cerebras API call by @marscod in #8862
• docs: update sambanova docs by @jhpiedrahitao in #8875
• Update model settings data by @yurchik11 in #8871
• (security fix) - Enforce model access restrictions on Azure OpenAI route by @krrishdholakia in #8888
• Show 'user_email' on key table on UI by @krrishdholakia in #8887
• fix: ollama chat async stream error propagation by @Tomas2D in #8870
• Litellm dev 02 27 2025 p6 by @krrishdholakia in #8891
• Fix calling claude via invoke route + response_format support for claude on invoke route by @krrishdholakia in #8908
• converse_transformation: pass 'description' if set in response_format by @krrishdholakia in #8907
• Fix bedrock passing response_format: {"type": "text"} by @krrishdholakia in #8900
• (Feat) - Show Error Logs on LiteLLM UI by @ishaan-jaff in #8904
• UI - Allow admin to control default model access for internal users by @krrishdholakia in #8912
• (bug fix - patch) - don't overwrite litellm.anthropic_models when running auth checks by @ishaan-jaff in #8924
• (patch) ui remove search button on internal users tab by @ishaan-jaff in #8926
• (bug fix) - don't log messages, prompt, input in model_parameters in StandardLoggingPayload by @ishaan-jaff in #8923
• Litellm stable release notes v1 61 20 by @krrishdholakia in #8929
• (bug fix) - dd tracer, only send traces when user opts into sending dd-trace by @ishaan-jaff in #8928
• docs(index.md): add demo instance to docs for easy testing by @krrishdholakia in #8931
• (Bug fix) - don't log messages in model_parameters in StandardLoggingPayload by @ishaan-jaff in #8932
• (UI) Error Logs improvements - Store Raw proxy server request for success and failure by @ishaan-jaff in #8917
• (UI) - Allow Internal Users to View their own logs by @ishaan-jaff in #8933
• Add supports_pdf_input: true for specific Bedrock Claude models by @iwamot in #8655
• Fix prometheus metrics w/ custom metrics + Handle sending image_url as str to openai by @krrishdholakia in #8935
• fix(proxy_server.py): fix setting router redis cache, if cache enable… by @krrishdholakia in #8859
• Fix relation "dailytagspend" does not exist error by @Schnitzel in #8947

New Contributors

• @marscod made their first contribution in #8862
• @jhpiedrahitao made their first contribution in #8875
• @Tomas2D made their first contribution in #8870
• @Schnitzel made their first contribution in #8947

Full Changelog: v1.61.20.rc...v1.62.1-nightly

Docker Run LiteLLM Proxy

docker run \
-e STORE_MODEL_IN_DB=True \
-p 4000:4000 \
ghcr.io/berriai/litellm:main-v1.62.1-nightly

Don't want to maintain your internal proxy? get in touch 🎉

Hosted Proxy Alpha: https://calendly.com/d/4mp-gd3-k5k/litellm-1-1-onboarding-chat

Load Test LiteLLM Proxy Results