Kaggle SimpleQA Verified: leaderboard
Google DeepMind's SimpleQA Verified, a 1,000-prompt revision of OpenAI's SimpleQA that removes noisy or incorrect labels, redundant questions and topical bias, from its official Kaggle leaderboard. Kaggle's task score is shown as a percentage (0-100, higher is better); the paper's headline metric is an F1 over correct and attempted answers.
Metric: Score (%). Source: www.kaggle.com. Saturation forecast: Around February 2027. 54 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 77.52 | #64 |
| 2 | GPT-6 Astra | 75.81 | #6 |
| 3 | Gemini 4 Argon | 75.6 | #9 |
| 4 | Gemini 3.8 Flash | 74.59 | #30 |
| 5 | GPT-6.1 Sol | 73.77 | #7 |
| 6 | Gemini 3.7 Flash | 73.19 | #24 |
| 7 | Gemini 3 Pro (Preview) | 72.1 | #71 |
| 8 | Gemini 3.5 Flash | 70.41 | #66 |
| 9 | GPT-5.6 Sol | 69.24 | #22 |
| 10 | GPT-6 Sol | 64.36 | #19 |
| 11 | Claude Opus 5 | 63.3 | #16 |
| 12 | Grok 4.7 | 59.37 | #26 |
| 13 | Grok 4.5 | 53.81 | #54 |
| 14 | Qwen 3 235B A22B 2507 (Thinking) | 53.72 | #263 |
| 15 | Grok 4.6 | 52.86 | #29 |
No result here: #3 Claude Opus 5.5, #10 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=kaggle-simpleqa-verified · How It Works · Data refreshed daily, snapshot 2026-11-05.