Creator
OpenAI
Released
2025-01-31
Intelligence
25.2
Artificial Analysis Index
Coding
17.3
Artificial Analysis Index
In $/1M
$1.10
input tokens
Out $/1M
$4.40
output tokens
Blended $/1M
$1.93
3:1 blended
Speed
197
tokens / sec
Benchmark breakdown
Independent evaluation scores, normalized to 0–100.
GPQA Diamond
77
Humanity’s Last Exam
12
MMLU-Pro
80
SciCode
40
LiveCodeBench
73
MATH-500
98
AIME 2025—
τ²-Bench (agentic)
31
Terminal-Bench Hard
6
IFBench
67