v1.87.0
📦 litellmView on GitHub →
✨ 8 features🐛 23 fixes🔧 19 symbols
Summary
This release focuses heavily on stability, bug fixes across various providers (Bedrock, Cohere, Deepseek, Vertex AI), and introduces new features like Gemini 3.5 Flash support and enhanced UI/Proxy capabilities. It also mandates new steps for verifying Docker image authenticity via cosign.
Migration Steps
- Users running Docker images should verify image signatures using the provided cosign commands (pinned commit hash recommended).
✨ New Features
- Added support for Gemini 3.5 Flash.
- Added support for Gemini managed agents.
- Added gemini-3.1-flash-lite model cost map.
- Propagate team_id and team_alias to all child OTEL spans.
- Added Interactions API endpoint to playground with SSE streaming.
- Added native MCP OAuth support for cursor.
- Migrated Interactions to Google Interactions API steps schema (May 2026).
- Add granian as an ASGI compliant web server for better throughput stability.
🐛 Bug Fixes
- Proxy admins can now gate team allowed_passthrough_routes.
- Stabilized image-edit VCR cassettes to stop live gpt-image-1 spend.
- Fixed Bedrock/Cohere to send embedding_types as JSON array, not string.
- Migrated realtime + rerank tests off shut-down upstream models.
- Fixed caching to replay openai/responses bridge cache hits as chat streams.
- Sanitized batch metadata in Bedrock to prevent Pydantic ValidationError.
- Fixed Deepseek to use native /anthropic/v1/messages endpoint and sanitize tools.
- Fixed proxy to decode bytes and pass-through SSE for Google-native streamGenerateContent.
- Replaced shut-down gpt-4o-audio-preview with gpt-audio-1.5 in tests.
- Fixed spend_counter to seed Redis counter via SET NX to prevent cross-pod double-seed.
- Normalized batch file IDs before ManagedObjectTable write in proxy.
- Fixed router to use forwarded model_id for native Azure container IDs.
- Restored log filter loading indicator in UI.
- Fixed JWT on tools/list and REST tools/call server resolution for MCP.
- Fixed Vertex AI to omit function_call id on Vertex Gemini 3.5+ tool turns.
- Fixed interactions to never drop streamed text deltas; always emit terminal completion.
- Exposed Prisma idle/connect timeout + extra DB URL params in proxy.
- Serialized guardrail_response to JSON in OTEL traces.
- Hydrated wildcard discovery credentials in proxy.
- Fixed Vertex Gemma to strip `context_management` from request body.
- Recalculated cost after router retry failures.
- Fixed OTEL to emit guardrail span on violation, surface status + categories.
- Fixed Docker to restore npm to non_root builder image.