NonoBench (v1.1): leaderboard
Metric: Standard accuracy (%), correct / non-failed attempts; version 1.1. Source: www.nonobench.com. Saturation forecast: Around January 2027. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) (High) | 76.67 |
| 2 | GPT-5.4 (xHigh) | 76.67 |
| 3 | GPT-5.4 (High) | 66.67 |
| 4 | GPT-5.4 (Low) | 50 |
| 5 | Gemini 3.1 Pro (Preview) (Low) | 43.33 |
| 6 | MiniMax-M2.5 | 26.67 |
| 7 | GLM-5 (Thinking) | 23.33 |
| 8 | GLM-5 (Non-reasoning) | 0 |
Interactive version: theaggregate.ai/benchmark?slug=nonobench-v1-1 · How It Works · Data refreshed daily, snapshot 2026-10-09.