gpqa
Unknown
OCR benchmark
17
Total Results
17
Models Tested
1
Metrics
2026-03-06
Last Updated
accuracy
Higher is better
| Rank | Model | Score | Source |
|---|---|---|---|
| 1 | o3 | 82.8 | openai-simple-evals |
| 2 | o4-mini | 77.6 | openai-simple-evals |
| 3 | o1 | 75.7 | openai-simple-evals |
| 4 | o3-mini | 74.9 | openai-simple-evals |
| 5 | o1-preview | 73.3 | openai-simple-evals |
| 6 | gpt-45-preview | 69.5 | openai-simple-evals |
| 7 | gpt-41 | 66.3 | openai-simple-evals |
| 8 | o1-mini | 60 | openai-simple-evals |
| 9 | claude-35-sonnet | 59.4 | openai-simple-evals |
| 10 | grok-2 | 56 | openai-simple-evals |
| 11 | llama-31-405b | 50.7 | openai-simple-evals |
| 12 | claude-3-opus | 50.4 | openai-simple-evals |
| 13 | gpt-4o | 49.9 | openai-simple-evals |
| 14 | gpt-4-turbo | 49.3 | openai-simple-evals |
| 15 | gemini-15-pro | 46.2 | openai-simple-evals |
| 16 | llama-31-70b | 41.7 | openai-simple-evals |
| 17 | gpt-4o-mini | 40.2 | openai-simple-evals |