LLM application teams that need automated prompt testing and production observability.

Athina AI, formerly athina-evals, is a testing and observability platform that runs automated evals against LLM outputs and traces production requests.

Available as a hosted cloud platform with an open source Python and TypeScript SDK on GitHub.

Launched around 2023 and actively maintained into 2026 under the athina-ai GitHub organization.

Helps teams catch regressions and hallucinations before they reach users and monitor cost, latency and quality in production.

Integrates with OpenAI, Anthropic and other LLM providers via API keys and a Python or TypeScript SDK.

SDK is open source, repo moved to athina-ai/athina-evals, but the full platform is a commercial hosted service requiring an account and API key, check what trace data leaves your infrastructure.