Langfuse

Langfuse is the leading open-source LLMOps platform, designed to help teams collaboratively develop, monitor, evaluate, and debug AI applications. With Langfuse, teams can self-host the platform in just a few minutes. It’s battle-tested, reliable, and trusted by thousands of users—from YC startups to major companies like Khan Academy and Twilio—who use it in production every day. Developers can trace any Large Language Model (LLM) or framework using our Python and JS/TS SDKs, open API, or native integrations with platforms like OpenAI, Langchain, Llama-Index, and the Vercel AI SDK. Beyond tracing, Langfuse offers powerful tools like Langfuse Prompt Management, open APIs, and testing and evaluation pipelines to enhance the quality of AI applications. Product managers can analyze, evaluate, and debug AI products with detailed metrics on costs, latencies, and user feedback, all available through the Langfuse Dashboard. They can also engage human annotators with customized workflows to score and improve the application. Langfuse provides a full suite of tools for managing security risks, including security frameworks and evaluation pipelines to ensure robust, secure AI deployments.