Guesswork 2026-06: leaderboard
Guesswork standings for 2026-06: each contestant predicts newly scraped (model, benchmark) scores from day-before information only, scored as mean absolute error in z-score units — lower is better. June was the immature-matrix month, when several LLMs beat the aggregate.
Metric: MAE (z-score units). Source: theaggregate.ai. Saturation forecast: Around September 2027. 19 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 5 | 0.72 | #12 |
| 2 | GPT-5.5 | 0.76 | #31 |
| 3 | Grok 4.6 | 0.79 | #26 |
| 4 | Claude Sonnet 5 | 0.79 | #57 |
| 5 | GPT-5.1 | 0.84 | #135 |
| 6 | Grok 4.3 | 0.9 | #188 |
| 7 | Kimi K2.6 | 0.91 | #104 |
| 8 | GLM-5.2 | 0.94 | #73 |
| 9 | Qwen 3.5 9B | 0.98 | #379 |
| 10 | Mistral Small 3 | 1 | #627 |
| 11 | Ling-2.6-flash | 1.01 | #532 |
| 12 | Gemini 2.5 Pro | 1.03 | #152 |
| 13 | DeepSeek V4 Pro | 1.08 | #102 |
| 14 | GPT-OSS-20B | 1.23 | #518 |
| 15 | Mistral Nemo | 1.41 | #1085 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=guesswork-2026-06 · How It Works · Data refreshed daily, snapshot 2026-10-20.