New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
Omni-MATH-HARD benchmark
AI model leaderboard for the Omni-MATH-HARD benchmark. Compare how large language models score on Omni-MATH-HARD, see the full ranking, and understand what this AI benchmark measures. GPT-5 currently leads with 73.61%. 173-problem hardest subset of Omni-MATH at olympiad-grade difficulty.
Leaderboard
| # | Model | Organization | Score | Variant | Source |
|---|---|---|---|---|---|
| #1 | GPT-5 | OpenAI | 73.61% | High | official ↗ |
| #2 | gemini-2.5-pro | Google DeepMind | 69.36% | official ↗ | |
| #3 | K2 Think | Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) | 60.73% | avg@16 | official ↗ |
| #4 | o3 | OpenAI | 59.39% | High | official ↗ |
| #5 | OpenReasoning-Nemotron-32B | NVIDIA | 58.88% | official ↗ | |
| #6 | gpt-oss-120b | OpenAI | 57.76% | official ↗ | |
| #7 | Qwen3-235B-A22B | Qwen | 56.91% | official ↗ | |
| #8 | DeepSeek-V3.1 | DeepSeek | 53.22% | official ↗ | |
| #9 | DeepSeek-R1 | DeepSeek | 51.33% | official ↗ | |
| #10 | QwQ-32B | Qwen | 46.93% | official ↗ | |
| #11 | gpt-oss-20b | OpenAI | 41.51% | official ↗ | |
| #12 | Qwen3-30B-A3B | Qwen | 23.87% | official ↗ |
Frequently asked questions about Omni-MATH-HARD
What is the Omni-MATH-HARD benchmark?
173-problem hardest subset of Omni-MATH at olympiad-grade difficulty.
How is the Omni-MATH-HARD benchmark scored?
Omni-MATH-HARD is scored using the accuracy (%) metric, where a higher score is better. The maximum achievable score is 100.000. GenAIList aggregates reported scores from model providers and papers into a single ranked leaderboard.
Which AI model scores highest on Omni-MATH-HARD?
As of the latest reported scores on GenAIList, GPT-5 achieves the highest result on Omni-MATH-HARD with a score of 73.61%.
Is a higher Omni-MATH-HARD score better?
Yes. On Omni-MATH-HARD a higher score indicates better performance, so models near the top of the leaderboard are the strongest.