Langfuse
Trace, debug, and evaluate your LLM application in production. Langfuse shows you exactly which prompts are failing, what they cost, and where quality is slipping.
Overview
Langfuse is an open-source LLM observability and evaluation platform — it traces every call your AI application makes, measures quality, tracks costs, and surfaces issues before they reach users. Building a production LLM app without observability is building blind: you don't know which prompts are failing, where latency is coming from, or whether the model is drifting. Langfuse wraps your existing LLM calls with a single integration and gives you a full dashboard of traces, scores, user sessions, and cost breakdowns. Teams use it to debug misbehaving agents, compare prompt versions, run evals on output quality, and track how model changes affect user experience over time.
Key Features
- Full LLM call tracing
- Prompt version management
- User session tracking
- Cost and latency analytics
- Evaluation datasets
- Self-hostable
- • One of the best open-source options in LLM observability
- • Works with any LLM provider
- • Eval framework helps catch quality regressions early
- • Setup requires SDK integration in your codebase
- • Dashboard can feel complex for simple use cases