PerceptionBench - Comparison — leaderboard
Metric: Comparison (%). Source: huggingface.co. 16 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.6 Sol | 67 |
| 2 | GPT-5.5 | 65.6 |
| 3 | Gemini 3.1 Pro (Preview) | 61.7 |
| 4 | Seed 2.1 Pro | 59.5 |
| 5 | Kimi K3 | 59.1 |
| 6 | Claude Fable 5 | 56.1 |
| 7 | Qwen 3.7 Plus | 55.9 |
| 8 | Gemini 3.5 Flash | 54.8 |
| 9 | Qwen 3.5 397B A17B | 49.8 |
| 10 | Claude Opus 4.8 | 48.4 |
| 11 | GLM-5V Turbo | 45.2 |
| 12 | Gemma 4 31B | 43.7 |
| 13 | Grok 4.5 | 43.7 |
| 14 | Kimi K2.6 | 39.1 |
| 15 | GLM-4.6V | 34.8 |
Interactive version: theaggregate.ai/benchmark?slug=perceptionbench-comparison · How It Works · Data refreshed daily, snapshot 2026-08-05.