GPT-6 Astra
OpenAI#1 of 2 by best run
Best success72%in Codex
Cheapest run$1.20per task in OpenCode
Fastest run122sper task in Hermes Agent
- Best harness
- Codex
- Graded runs
- 6 of 7 harnesses
- Latest run
- 2026-09-10
Success rate
72.4%
Hermes Agent72.4%
72.4%
69.0%
65.5%
65.5%
Time per task
Hermes Agent122s
126s
127s
151s
159s
194s
Runs
#RunSuccess$ / taskTime / task
- 01
Codex
72%$ / task$1.63Time / task151s - 02
Hermes Agent72%$ / task$1.51Time / task122s - 03
Command Code
72%$ / task$2.06Time / task159s - 04
Claude Code
69%$ / task$15.72Time / task194s - 05
OpenCode
66%$ / task$1.20Time / task126s - 06
Pi Agent
66%$ / task$1.29Time / task127s
1 harness not yet run. Pages fill in as new runs land