Benchmarks · knowledge
SimpleQA Verified (Epoch AI run)
Google DeepMind's cleaned-up SimpleQA short-answer factuality set, run by Epoch AI. Share answered correctly. Data: Epoch AI, CC BY.
82 models · 89 results · data as of 2026-10-29 · epoch.ai ↗
Leaderboard
Gemini 4 Argon high
All results
89 rows| # | Model | Effort | Score | Harness |
|---|---|---|---|---|
| 1 | Gemini 4 Argon | high | 76.0% | Epoch AI |
| 2 | GPT-6 Astra | max | 75.6% | Epoch AI |
| 3 | GPT-6.1 Sol | max | 73.9% | Epoch AI |
| 4 | Gemini 3.1 Pro | high | 73.5% | Epoch AI |
| 5 | Opus 5.5 | max | 72.2% | Epoch AI |
| 6 | Fable 5.1 | max | 70.8% | Epoch AI |
| 7 | Fable 5 | xhigh | 70.7% | Epoch AI |
| 8 | Gemini 3.8 Flash | high | 69.7% | Epoch AI |
| 9 | GPT-5.6 Sol | max | 69.7% | Epoch AI |
| 10 | Gemini 3.7 Flash | high | 69.2% | Epoch AI |
| 11 | Gemini 3 Flash | high | 66.8% | Epoch AI |
| 12 | Gemini 3.5 Flash | high | 66.2% | Epoch AI |
| 13 | Gemini 3.6 Flash | high | 66.2% | Epoch AI |
| 14 | GPT-5.5 | xhigh | 63.0% | Epoch AI |
| 15 | GPT-6 Sol | max | 60.7% | Epoch AI |
| 16 | Muse Spark 1.2 | xhigh | 60.3% | Epoch AI |
| 17 | Opus 5 | max | 59.9% | Epoch AI |
| 18 | Muse Spark 1.1 | max | 57.8% | Epoch AI |
| 19 | GPT-6.1 Terra | max | 57.3% | Epoch AI |
| 20 | Grok 4.7 | xhigh | 56.0% | Epoch AI |
| 21 | Qwen 3.7 Max | max | 55.8% | Epoch AI |
| 22 | Opus 4.8 | max | 53.0% | Epoch AI |
| 23 | DeepSeek V4 Pro 0813 | max | 52.9% | Epoch AI |
| 24 | Qwen 3.6 Max | max | 52.0% | Epoch AI |
| 25 | Opus 4.7 | xhigh | 51.7% | Epoch AI |
| 26 | Kimi K3 | max | 50.6% | Epoch AI |
| 27 | GPT-5 | high | 50.1% | Epoch AI |
| 28 | o3 | high | 49.4% | Epoch AI |
| 29 | Grok 4.6 | high | 49.3% | Epoch AI |
| 30 | Grok 4.6 | xhigh | 48.9% | Epoch AI |