Langfuse

Langfuse is an open-source LLM engineering platform designed to help developers build, debug, and optimize large language model applications. Key features include:

Tracing: Gain detailed insights into your LLM app's behavior with production-level tracing, allowing for faster debugging and performance optimization.

Prompt Management: Collaborate on prompt versioning and deployment, with low-latency retrieval for seamless integration into your workflow.

Evaluation: Collect user feedback, perform annotations, and run evaluation functions directly within Langfuse to assess and improve your LLM's output quality.

Metrics: Track crucial performance indicators such as cost, latency, and quality to ensure your LLM application meets both technical and business requirements.

Integrations: Langfuse offers SDKs for Python and JavaScript/TypeScript, along with native integrations for popular LLM libraries and frameworks like LangChain, OpenAI, and Llama-Index.

Open Source and Enterprise-Ready: Fully open-source and can be self-hosted, while also offering a cloud version with enterprise-grade security (SOC 2 Type II and ISO 27001 certified, GDPR compliant).

With over 2.9 million SDK installs per month and 6,995+ GitHub stars, Langfuse is trusted by teams at companies like PostHog, Samsara, and Khan Academy for building complex LLM applications. Whether you're developing a proof of concept or scaling to production, Langfuse provides the tools necessary to engineer high-quality LLM-powered solutions.