langfuse

🪢 Open source AI engineering platform: LLM evals, observability, metrics, prompt management, playground, datasets. Integrates with OpenTelemetry, LangChain, OpenAI SDK, LiteLLM, and more. 🍊YC W23

GitHub langfuse.com

analyticsautogenevaluationlangchainlarge-language-modelsllama-indexllmllm-evaluationllm-observabilityllmopsmonitoringobservabilityopenaiopen-sourceplaygroundprompt-engineeringprompt-managementself-hostedycombinator

Browse cluster: Claude AI Agent Frameworks & MCP Tools →

8,473commits

203contributors

9languages

Tech stack & purpose

Langfuse is an open source LLM engineering platform built with Next.js for the web frontend, a separate worker service for background processing, and shared domain logic in TypeScript packages. The platform provides LLM observability, evaluation, metrics, prompt management, a playground, and dataset management capabilities, integrating with OpenTelemetry, LangChain, OpenAI SDK, LiteLLM, and other tools. The project is developed by a team backed by Y Combinator (W23) and includes enterprise features such as Stripe-based billing with flexible usage-based pricing, subscription management, and cloud infrastructure support. The codebase is organized as a monorepo with a Next.js web application, background worker, shared Postgres and ClickHouse databases, and uses OpenTelemetry throughout for observability and metrics collection.