Guesswork 2026-07: leaderboard
Metric: MAE (z-score units). Source: theaggregate.ai. 37 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GLM-5.2 | 0.67 |
| 2 | Claude Opus 5 | 0.68 |
| 3 | Grok 4.6 | 0.73 |
| 4 | Gemini 3.1 Pro (Preview) | 0.77 |
| 5 | Claude Sonnet 5 | 0.81 |
| 6 | Gemini 2.5 Pro | 0.82 |
| 7 | DeepSeek V4 Pro | 0.82 |
| 8 | Qwen 3.7 Max | 0.87 |
| 9 | Grok 4.3 | 0.91 |
| 10 | Mistral Small 4 | 0.93 |
| 11 | Kimi K2.6 | 0.98 |
| 12 | GPT-5.5 | 0.98 |
| 13 | codestral-2508 | 1 |
| 14 | Claude Opus 4.8 | 1 |
| 15 | Mistral Medium 3.1 | 1.01 |
Interactive version: theaggregate.ai/benchmark?slug=guesswork-2026-07 · How It Works · Data refreshed daily, snapshot 2026-09-05.