Langfuse - Open Source Agent Evals & Observability
Langfuse is one integrated, open platform to launch, observe, and improve AI agents. Use production data to understand agent behavior, collaborate on fixes, and ship better quality at lower cost and latency.
Key Features
- Tracing and observability: See what your LLM apps and agents are doing in production.
- Evaluations: Evaluate and improve agent quality across the full development cycle.
- Open source (MIT): Self-host or use the managed cloud.
- OpenTelemetry native: Works with any stack and 100+ integrations.
- Built for scale: Async by default with enterprise-grade security.
Perfect for: Developers and teams building LLM applications and AI agents.
