o1-pro
OpenAI🇺🇸 United States
o1 with extended compute for the hardest problems. Significantly outperforms o1 on math, science, and complex coding tasks.
Context window200K tokens
Input / 1M tokens$150
Output / 1M tokens$600
Version History
o1-pro-2024-12-17major
o1-pro launches with extended compute for the hardest reasoning tasks. Sets new records on competition math and frontier science benchmarks.
Benchmark Scores
Full leaderboard →86.7%
AIME 2024
85.0%
AIME 2025
79.7%
GPQA
91.0%
HumanEval
93.8%
MATH
92.3%
MMLU