●Change8

v1.80.10.rc.1

📦 litellm
✨ 22 features🐛 25 fixes🔧 17 symbols

Summary

LiteLLM v1.80.10 introduces Azure GPT‑5.2 models, new security guardrail evidence, expanded UI features, and numerous bug fixes and documentation updates.

✨ New Features

  • Add Azure GPT‑5.2 models
  • Add opt‑in evidence results for Pillar Security guardrail during monitoring
  • Add support for \"expires after\" parameter in Files endpoint
  • Add support for forwarding client headers in /rerank endpoint
  • Add v0 support for target storage
  • Add new Badge for Agent Usage
  • JWT Auth – allow selecting team_id from request header
  • Add latency metrics (TTFT, TPOT, Total Generation Time) to OTEL payload
  • Add benchmark_proxy_vs_provider.py script
  • UI – show UI version on top left near logo
  • OpenAI – add reasoning_effort='xhigh' support for GPT‑5.2 models
  • UI playground – allow custom model name as option[0]
  • Add support for using MCPs on /chat/completions
  • UI – reorganize left navigation categories and display agents on root
  • Add Usage Entity labels
  • Bedrock – add EU Claude Opus 4.5 model
  • Add agent cost tracking on UI
  • A2A Gateway – allow adding Azure Foundry Agents via UI
  • Show progress and pause on hover for Notifications
  • Add documentation for target storage
  • Add health endpoint tests to CI with database and Redis support
  • Add benchmark script for proxy vs provider performance

🐛 Bug Fixes

  • Add semver prerelease suffix to Helm chart versions
  • Add speechConfig to GenerationConfig for Gemini TTS
  • Prometheus logger – add 'exception_status' field
  • Cache – handle string content in is_cached_message
  • Embeddings – fix handling token array input decoding
  • Router – add minimum request threshold for error‑rate cooldown
  • Fix Milvus client documentation
  • Use Docker executor for CI jobs
  • Remove dependency on database and Redis from health test
  • Router – replace time.perf_counter() with time.time()
  • Fix x‑litellm‑key‑spend header update
  • Bedrock – fix header forwarding with custom API
  • Add all proxy models to default user settings
  • UI – minor issues on Agent Usage page
  • Fix MCP tool name prefix
  • CI/CD – fix mypy, code quality, and MCP testing checks
  • CI/CD – fix failing proxy unit test and Langfuse trace_id test
  • CI/CD – fix failing proxy and core integration tests
  • CI/CD – fix JWT authentication tests failing with KeyError: 'headers'
  • UI – display request and response in logs
  • Regex guardrails update
  • CI/CD – fix Bedrock tool calling test failures with non‑serializable objects and internal parameters
  • Filter internal parameters in fallback code
  • UpdateMCPServerRequest – add extra_headers and allowed_tools
  • A2A Gateway – ensure Azure Foundry agents work correctly

🔧 Affected Symbols

GenerationConfigprometheus loggeris_cached_messageembeddings token array decoderrouter error‑rate cooldown logicschema.prisma (storage_backend, storage_url columns)UpdateMCPServerRequestGCS PubSub spend_logs_payload.json (agent_id field)OTEL payload (latency metrics)MCP handling in /chat/completionsFiles endpoint (expires after param)/rerank endpoint (client header forwarding)target storage moduleUI components (badge, version display, left navigation, notifications)JWT authentication module (team_id selection)A2A gateway Azure Foundry agent integrationregex guardrails module