SimpleQA Verified
Rank #4 of 51 · 51 models scored · top score 77.3
unknown
Strongest on discrimination, weakest on saturation headroom. Caveat: the public test set leaves contamination on the table.
Weighted composite. Each signal is computed from the source-backed leaderboard below.
Does it still separate models?
How far from ceiling / clustered at the top?
Public vs held-out; training-leak risk.
Apples-to-apples, or mixed / vendor-optimized?
How old is the benchmark?
| # | Model | Score | Evidence | Measured |
|---|---|---|---|---|
| 1 | Gemini 3.1 Pro | 77.3 | reproduced | 2026-02-19 |
| 2 | Gemini 3 Pro | 72.9 | reproduced | 2025-11-18 |
| 3 | Gemini 3.5 Flash | 68.4 | reproduced | 2026-05-19 |
| 4 | Claude Fable 5 | 68.3 | reproduced | 2026-06-09 |
| 5 | Qwen3-Max | 67.47 | reproduced | 2025-09-05 |
| 6 | Gemini 3 Flash | 67.4 | reproduced | 2025-12-17 |
| 7 | Muse Spark | 66.3 | reproduced | 2026-04-08 |
| 8 | GPT-5.5 Pro | 64.5 | reproduced | 2026-04-23 |
| 9 | GPT-5.5 | 63.1 | reproduced | 2026-04-23 |
| 10 | Qwen3.7-Max | 58.52 | reproduced | 2026-05-19 |
| 11 | DeepSeek-V4-Pro | 57 | reproduced | 2026-04-24 |
| 12 | Qwen 3.6 Max (Preview) | 56.93 | reproduced | 2026-04-20 |
| 13 | Gemini 2.5 Pro (Jun 2025) | 56 | reproduced | 2025-06-05 |
| 14 | o3 | 53 | reproduced | 2024-12-20 |
| 15 | Claude Opus 4.7 | 50.6 | reproduced | 2026-04-16 |
| 16 | GPT-5 | 50.6 | reproduced | 2025-08-07 |
| 17 | Qwen3-235B-A22B-Thinking (Jul 2025) | 50.1 | reproduced | 2025-07-25 |
| 18 | Qwen 3.6 Plus | 49.1 | reproduced | 2026-04-01 |
| 19 | GPT-5.1 | 48.9 | reproduced | 2025-11-13 |
| 20 | Grok 4 | 47.9 | reproduced | 2025-07-09 |
| 21 | GPT-5.4 Pro | 47.8 | reproduced | 2026-03-05 |
| 22 | Claude Opus 4.6 | 46.49 | reproduced | 2026-02-05 |
| 23 | GPT-5.4 | 44.83 | reproduced | 2026-03-05 |
| 24 | Claude Opus 4.5 | 41.8 | reproduced | 2025-11-24 |
| 25 | Claude Opus 4.8 | 39.5 | reproduced | 2026-05-28 |
| 26 | Kimi K2.7 Code | 39.2 | reproduced | 2026-06-12 |
| 27 | GPT-5.2 | 38.9 | reproduced | 2025-12-11 |
| 28 | Kimi K2.6 | 38.7 | reproduced | 2026-04-20 |
| 29 | GLM-5.2 | 38.1 | reproduced | 2026-06-16 |
| 30 | Grok 4.3 Beta | 38 | reproduced | 2026-04-17 |
| 31 | GLM-5.1 | 37.29 | reproduced | 2026-04-07 |
| 32 | Claude Opus 4.1 | 34.8 | reproduced | 2025-08-05 |
| 33 | Kimi K2.5 | 33.9 | reproduced | 2026-02-02 |
| 34 | Kimi K2 Thinking | 31.6 | reproduced | 2025-11-06 |
| 35 | GLM-4.7 | 31.5 | reproduced | 2025-12-22 |
| 36 | Claude Sonnet 4.6 | 29 | reproduced | 2026-02-17 |
| 37 | GPT-5.4 Mini | 28.6 | reproduced | 2026-03-17 |
| 38 | DeepSeek-V3.2 | 27.5 | reproduced | 2025-12-01 |
| 39 | DeepSeek-R1 (May 2025) | 27.4 | reproduced | 2025-05-28 |
| 40 | Qwen 3.5 Plus (hosted 397B-A17B) | 26 | reproduced | 2026-02-16 |
| 41 | o4-mini | 23.9 | reproduced | 2025-04-16 |
| 42 | Claude Sonnet 4.5 | 23.6 | reproduced | 2025-09-29 |
| 43 | Qwen 3.6 Flash | 21.16 | reproduced | 2026-04-26 |
| 44 | Grok-3 mini | 21.1 | reproduced | 2025-02-19 |
| 45 | GPT-5 mini | 21 | reproduced | 2025-08-07 |
| 46 | Qwen 3.5 Flash (hosted 35B-A3B) | 19.76 | reproduced | 2026-02-25 |
| 47 | gpt-oss-120b | 13.9 | reproduced | 2025-08-05 |
| 48 | GPT-5 nano | 12.2 | reproduced | 2025-08-07 |
| 49 | GPT-5.4 Nano | 12 | reproduced | 2026-03-17 |
| 50 | Claude 3.5 Haiku | 6.7 | reproduced | 2024-10-22 |
| 51 | Claude Haiku 4.5 | 5.9 | reproduced | 2025-10-15 |