v1.81.11-nightly
📦 litellmView on GitHub →
✨ 8 features🐛 23 fixes🔧 13 symbols
Summary
This release introduces new configuration options, enhanced guardrail capabilities, and support for new models like MiniMax. Numerous bug fixes address issues related to streaming errors, queue management, Azure compatibility, and model metadata synchronization.
Migration Steps
- If you rely on Python tracebacks being present in streaming SSE error responses, be aware they are now suppressed.
- If you were using the old 'HTTP' transport type in the Add New MCP Server page, it has been renamed to 'Streamable HTTP (Recommended)'.
✨ New Features
- Added standard_logging_payload_excluded_fields configuration option.
- Generic Guardrails now forward request headers and litellm_version.
- Organization and Team Admins can now call /invitation/new.
- Added support for sorting on /spend/logs/ui.
- Added toxic/abusive content filter guardrails.
- Added support for MiniMax-M2.1 and MiniMax-M2.1-lightining models.
- Added support for server-side compaction on the OpenAI Responses API context_management.
- Added Tracing for MCP Calls running through AI Gateway.
🐛 Bug Fixes
- Handled metadata=None in SDK path retry/error logic (utils.py).
- Stopped leaking Python tracebacks in streaming SSE error responses.
- Used anyio.fail_after instead of asyncio.wait_for for StreamableHTTP backends.
- Removed orphan entries from the scheduler queue, resolving a memory leak.
- Avoided in-place mutation in SpendUpdateQueue aggregation.
- Used current retry exception for retry backoff in the router.
- reasoning_effort=None and "none" now correctly return None for Opus 4.6.
- Preserved content_policy_violation error details from Azure OpenAI.
- Fixed OCI Cohere system messages by populating preambleOverride.
- Synchronized DeepSeek model metadata and added bare-name fallback.
- Exported PermissionDeniedError from litellm.__init__.
- Removed x-anthropic-billing block.
- Enabled verbose_logger when LITELLM_LOG=DEBUG.
- Merged query params when authorization_url already contains them in MCP.
- Fixed Gemini multi-turn tool calling message formatting (Note: This fix was reverted later).
- Supported Azure AD token auth for non-Claude azure_ai models.
- Supported prompt_cache_key for OpenAI and Azure chat completions.
- Fixed UI - Guardrail Edit: LiteLLM Content Filter Categories.
- Fixed UI - Add Auto Router: Description Text Input Focus.
- Added claude opus 4.6 in _supports_tool_search_on_bedrock.
- Reverted fix for Gemini multi-turn tool calling message formatting (PR #21051).
- Fixed authorization issues regarding same alias.
- Skipped premium check for empty metadata fields on team/key update.