ComposioBENCH
Try Bench

Codex

OpenAI#2 of 8 by coverage
Compare harnesses
Best success76%with GPT-6 Astra
Fastest run149sper task with GPT-6 Astra
Model support
Open
Graded runs
2 of 2 models
Latest run
2026-10-15

Success rate

% of tasks passed by model · higher is better
GPT-6 Astra
75.9%
DeepSeek V4 Pro
53.3%

Time per task

Seconds per task by model · lower is better
GPT-6 Astra
149s
DeepSeek V4 Pro
482s

Runs

  1. 01GPT-6 Astra
    76%
    Time / task149s
  2. 02DeepSeek V4 Pro
    53%
    Time / task482s