Agenta
Open-source LLMOps platform combining prompt management, evaluation, and observability for teams building production LLM apps.
About
Agenta is an open-source platform that covers the full lifecycle of developing and maintaining LLM-powered applications. Teams use its prompt playground to test and compare prompts against real test cases across more than 50 model providers, then version and promote configurations through branching and environment management. The evaluation module includes over 20 built-in evaluators plus support for custom scoring and human feedback. An observability layer built on OpenTelemetry tracks costs, latency, and detailed traces across every LLM call. Agenta can be self-hosted for free under the MIT license, and Agenta Cloud offers a free tier with no credit card required, alongside paid plans for larger teams.