o3
radar chart — all benchmarks
family — openai-o
benchmark scores Scores from Mar 2026
| Categories | ||
|---|---|---|
| MATH / AIME | 96.7% | reasoning |
| SimpleQA | 96.7% | general |
| MMLU | 92.9% | general |
| GPQA (Diamond) | 87.7% | reasoning |
| HumanEval / MBPP | 87.4% | coding |
| LiveCodeBench | 82.7% | coding |
| SWE-bench | 71.7% | agenticcoding |
| Chatbot Arena (LMSYS) | 1424 | general |
| TAU-bench | — | agenticmultiagent |
| GAIA | — | agenticmultiagent |
| WebArena | — | agentic |
| MT-Bench | — | general |
| AgentBench | — | multiagent |
| IFEval | — | generalagentic |
pricing — per 1M tokens via openrouter
Data unavailable
latency percentiles — time to first token (ms)
Data unavailable
model specifications
Data unavailable