Evals & Testing

Atla

Builds dedicated evaluator/judge models (Selene) and agent error-analysis tooling rather than a full observability suite.

Website ↗
CategoryEvals & Testing
Open sourceNo
Self-hostableNo
Pricing modelfreemium
Pricing notesAPI free tier; usage-based pricing
Framework integrationsopenai-sdk, langchain, crewai
Funding / ownership$5M seed (2025, incl. Y Combinator)

Pricing/feature source: https://www.atla-ai.com