Rank #775 · on radar since 2026-07-03

ariadne-eval

Trajectory-level observability and evaluation for LLM agents. Trace, score, and detect regressions in multi-step tool-using agent runs.

Why it's ranked

Every score decomposes into published factors — the same math for every tool, paid or not. Read the methodology →