Langfuse
Open-source observability & evaluation platform for LLM/agent applications — tracing, metrics, datasets, prompt versioning, UI, self-hostable.
77
Hero Score
Popularity
80
Performance
85
Ecosystem
100
Maturity
61
Dev Experience
57
⭐ 28,297 stars⬇ 3.9M downloads/wkFirst release: Jun 2023Last release: May 2026
Async Support: YesPlugin Extensions: Very highSpeed: FastDoc Quality: HighLearning Curve: Medium
Pros
- • Fully open-source (MIT license) with easy self-hosting via single Docker container
- • Framework-agnostic with integrations for LangChain, LlamaIndex, OpenAI SDK, LiteLLM, and more
- • Built on OpenTelemetry standard with comprehensive observability, prompt management, and evaluation tools
Cons
- • Self-hosting requires infrastructure management and operational expertise for production deployments
- • Community support for troubleshooting compared to dedicated enterprise support teams
- • UI can become complex when scaling to large numbers of traces and evaluation datasets