New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
SuperGPQA benchmark
AI model leaderboard for the SuperGPQA benchmark. Compare how large language models score on SuperGPQA, see the full ranking, and understand what this AI benchmark measures. Pangu Pro MoE currently leads with 54.8. Graduate-level cross-discipline benchmark spanning 285 disciplines — extension of GPQA.
Leaderboard
| # | Model | Organization | Score | Variant | Source |
|---|---|---|---|---|---|
| #1 | Pangu Pro MoE | Huawei | 54.8 | instruct | official ↗ |
| #2 | Claude 3.5 Sonnet | Anthropic | 48.2 | 1022-cited-mimo | official ↗ |
| #3 | Llama 4 Scout | Meta AI | 45 | cited-pangu-pro-moe | official ↗ |
| #4 | DeepSeek-R1-Distill-Qwen-14B | Qwen | 40.6 | cited-mimo | official ↗ |
| #5 | DeepSeek-R1-Distill-Qwen-14B | DeepSeek | 40.6 | cited-mimo | official ↗ |
| #6 | DeepSeek-R1-Distill-Qwen-7B | Qwen | 28.9 | cited-mimo | official ↗ |
| #7 | DeepSeek-R1-Distill-Qwen-7B | DeepSeek | 28.9 | cited-mimo | official ↗ |
Frequently asked questions about SuperGPQA
What is the SuperGPQA benchmark?
Graduate-level cross-discipline benchmark spanning 285 disciplines — extension of GPQA.
How is the SuperGPQA benchmark scored?
SuperGPQA is scored using the pass@1 metric, where a higher score is better. GenAIList aggregates reported scores from model providers and papers into a single ranked leaderboard.
Which AI model scores highest on SuperGPQA?
As of the latest reported scores on GenAIList, Pangu Pro MoE achieves the highest result on SuperGPQA with a score of 54.8.
Is a higher SuperGPQA score better?
Yes. On SuperGPQA a higher score indicates better performance, so models near the top of the leaderboard are the strongest.