New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
o3
About o3
o3 is an AI model developed by OpenAI, in the language category, released in 2024, made available as a proprietary (API-only) model.
On this page you'll find o3's full specifications, including 25 benchmark results. Review provider pricing and benchmark scores below, or compare o3 head-to-head with other language models.
Links
Community
No reports yet. Own this model on some hardware? Be the first to confirm it runs.
No comments yet. Start the discussion.
o3 benchmark scores▶
| Benchmark | Category | Score | Variant |
|---|---|---|---|
| BrowseComp | agentic | 49.7% | |
| BrowseComp | agentic | 49.7% | react-cited-tongyi |
| BrowseComp-Zh | agentic | 58.1 | react-cited-tongyi |
| TAU-bench Airline | agentic | 52% | |
| TAU-bench Retail | agentic | 70.4% | reported in GLM-4.5 paper |
| Terminal-Bench 2.0 | agentic | 30.2% | |
| WebWalkerQA | agentic | 71.7 | react-cited-tongyi |
| xbench-DeepSearch | agentic | 67 | react-cited-tongyi |
| LiveCodeBench | code | 78.4% | 2407-2501 |
| LiveCodeBench v5 | code | 73.3% | High |
| SWE-bench Verified | code | 69.1% | |
| SciCode | code | 41% | |
| MMLU-Pro | general | 85.3% | |
| Frames | long-context | 84 | react-cited-tongyi |
| AIME 2024 | math | 92.26% | High |
| AIME 2024 | math | 90.3% | |
| AIME 2025 | math | 86.58% | High |
| HMMT February 2025 | math | 80.8% | High |
| MATH-500 | math | 99.2% | |
| Omni-MATH-HARD | math | 59.39% | High |
| GPQA Diamond | reasoning | 82.7% | |
| Humanity's Last Exam | reasoning | 24.9% | react-cited-tongyi |
| Humanity's Last Exam | reasoning | 22.34% | High |
| Humanity's Last Exam | reasoning | 20% | |
| BFCL v3 | tool-use | 72.4% |
More models like o3
Frequently asked questions
What is o3? ▶
o3 is an AI model developed by OpenAI, in the language category, released in 2024. It is tracked on GenAIList with its specifications, benchmark scores and provider pricing.
Who created o3? ▶
o3 was developed by OpenAI and released in 2024.
Is o3 open source or proprietary? ▶
o3 is a proprietary model. It is accessed through an API rather than by downloading the weights.
How much does o3 cost? ▶
Pricing for o3 depends on the provider. See the providers table on this page for the latest API rates.
How does o3 perform on benchmarks? ▶
o3 is benchmarked across 25 evaluations on GenAIList, including BrowseComp (49.7%). See the full benchmark table below and compare it with other models.
No videos yet. Know a good one? Add it below.
Know a great video about o3?
Add YouTube Video
Add Tweet/Discussion
Report an Issue
Help us maintain accurate information in our database.
Suggest an Edit
Tell us what needs to be changed or corrected for