●Change8

v1.79.1.dev2

📦 litellmView on GitHub →
✨ 11 features🐛 13 fixes🔧 13 symbols

Summary

This release introduces significant enhancements to vector store support (Milvus, Azure AI), adds custom provider support for video endpoints, and resolves several critical bugs related to Pydantic warnings, AWS Bedrock, and Azure API handling. It also includes various UI improvements and fixes.

Migration Steps

  1. If you are using Pydantic v2.11+, ensure your environment is compatible, as memory accumulation issues related to deprecation warnings were resolved.

✨ New Features

  • Milvus vector store search support added.
  • Azure AI Vector Stores now support "virtual" indexes and vector store creation via the passthrough API.
  • Added support for `custom_llm_provider` for non-generation video endpoints.
  • LLM provider response headers are now added to the Responses API.
  • E2E Container API Support added.
  • Guardrail information is now changed to a list type.
  • Milvus Passthrough API support added, including create and read vector store support.
  • Support for streaming responses API in langfuse otel.
  • UI Feature: Delete Team Member with friction.
  • UI Feature: Add Model Existing Credentials Improvement.
  • UI Feature: Guardrail Info Page now shows PII Config.

🐛 Bug Fixes

  • Resolved memory accumulation caused by Pydantic 2.11+ deprecation warnings.
  • Removed minimum validation for cache control injection index.
  • Fixed empty assistant message handling in AWS Bedrock Converse API to prevent 400 Bad Request errors.
  • Fixed issue where Azure didn't accept extra body parameters.
  • Fixed Anthropic token counting for VertexAI.
  • Fixed regression in UI where Guardrail Entity Could not be selected and the entity was not displayed.
  • Stripped base64 in S3 operations.
  • Fixed dot notation support on UI SSO.
  • Fixed returning the diarized transcript when required in the request.
  • Fixed index field not being populated in streaming mode with n>1 and tool calls.
  • Updated perplexity cost tracking.
  • Fixed `image_config.aspect_ratio` not working for `gemini-2.5-flash-image`.
  • Passthrough Endpoints now support multi-part form data.

🔧 Affected Symbols

Passthrough EndpointsAzure AI Vector StoresAWS Bedrock Converse APIResponses APIAnthropic token countingVertexAIGuardrail EntityUI SSOlangfuse otelguardrail_informationMilvus Passthrough APIimage_config.aspect_ratiogemini-2.5-flash-image