Braintrust, founded in 2023 by Ankur Goyal, builds an AI evaluation and observability platform that lets teams test, trace, and score AI product behavior before and after it ships, addressing the problem that most AI failures show up in production rather than in a test suite because there's no established equivalent of unit testing for LLM-based features.
That evaluation-first focus, rather than logging and tracing alone, drove strong enough enterprise traction to bring in an $80 million Series B in February 2026, led by ICONIQ with Andreessen Horowitz, Greylock, and Elad Gil participating, positioning Braintrust as one of the better-capitalized platforms in the AI observability space.
Pricing runs a straightforward two-tier structure: a free tier to start, then $249 a month with tracing billed at $3 per GB, with no mid-tier option in between. For a team shipping an LLM-based feature that needs a real way to evaluate output quality before and after deployment, not only log what happened after the fact, Braintrust's evaluation-first platform addresses that directly.
