← All benchmarks
SandboxBench (targets of 22)
We haven't written a description for this benchmark yet; see the card notes below for how each maker ran it.
Your workloadRAG over docs · 12K in · 600 out · 50K requests/mo · 30% cached
- ModelScoreYour cost / mo
- 1GPT-6 Sol · OpenAIbest value1 targets
“Best value”: no other model here reports a higher score at a lower cost for your workload. Scores are self-reported by each maker with different settings, so treat close calls as ties.
How each maker reported it
- GPT-6 Sol Appendix A, p.120–154
The card reports safety and Preparedness evaluations, not general capability benchmarks. Cyber scores are at maximum reasoning effort; GPT-5.6 Sol cyber values are its maximum scores using more tokens. HealthBench scores are length-adjusted. SandboxBench is targets solved out of 22. Astra values only where the appendix states them.
Only each model's own maker's number is shown; scores a maker reports for rival models are left out.