GPT-5.2 Pro
radar chart — all benchmarks
family — gpt-5
benchmark scores Scores from Mar 2026
| Categories | ||
|---|---|---|
| SWE-bench | — | agenticcoding |
| HumanEval / MBPP | — | coding |
| MMLU | — | general |
| GPQA (Diamond) | — | reasoning |
| MATH / AIME | — | reasoning |
| TAU-bench | — | agenticmultiagent |
| GAIA | — | agenticmultiagent |
| WebArena | — | agentic |
| Chatbot Arena (LMSYS) | — | general |
| MT-Bench | — | general |
| LiveCodeBench | — | coding |
| AgentBench | — | multiagent |
| IFEval | — | generalagentic |
| SimpleQA | — | general |
pricing — per 1M tokens via openrouter
Data unavailable
latency percentiles — time to first token (ms)
Data unavailable
model specifications
Data unavailable