AI & Agents

Terminal-Bench

Last updated 2026-07-18 · benchmark measured 2026-08-09 — deterministic & reproducible

Terminal-Bench is a leaderboard benchmarking AI agents on terminal-based coding tasks.

Is Terminal-Bench production-ready?

Legit.Show scores Terminal-Bench 74 out of 100 — the simple average of its 7 measured frames. Legit.Show ran its deterministic 7-Frame production-readiness benchmark on Terminal-Bench (public-surface assessment), measured from the public surface with no LLM in the scoring path. Its strongest frame is Reliability; its weakest is Privacy. Every frame it averages is published with its evidence on the Legit.Show listing.

The 7 Frames

• Performance — 87/100
• Accessibility — 96/100
• Security — 55/100
• Privacy — 25/100
• Reliability — 100/100
• Standards — 92/100
• Discoverability — 60/100

What we measured

• Security headers present: HSTS, X-Content-Type-Options.
• No Content-Security-Policy.
• Served over HTTPS with a valid certificate.
• Real Lighthouse performance run — 1332 ms to first byte.
• Returns a proper 404 for unknown routes.
• 3 of 3 sampled routes reachable.
• No privacy policy found.
• Sets cookies / loads scripts with no consent prompt.

Who it's for

AI researchers · Agent developers · Model evaluators · Organizations building agents

Visit Terminal-Bench → · Alternatives to Terminal-Bench → · How this was measured →