SimpleQA Verified F1
General QA
The official Google and Kaggle SimpleQA Verified metric over 1,000 factoid questions. It is the harmonic mean of answer accuracy and accuracy conditional on attempted answers, reported on a 0-100 scale. This F1 aggregate is distinct from the percentage-correct metric used by Epoch and cannot be treated as 1,000 binary successes.
Top models (higher is better)
| Model | Score |
|---|---|
| Gemini 3 Pro Preview | 72.1 |
| Gemini 2.5 Pro | 54.5 |
| GPT-5.1 | 34.9 |
| Sonnet 4.5 | 29.3 |