OmniaBench — leaderboard
Metric: Overall Pass@1 (%). Source: scuuy.github.io. 22 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 5 | 58.54 |
| 2 | GPT-5.6 Sol | 57.14 |
| 3 | GLM-5.2 | 56.83 |
| 4 | GPT-5.5 | 56.52 |
| 5 | DeepSeek V4 Pro | 54.5 |
| 6 | Claude Opus 4.7 | 54.19 |
| 7 | Kimi K2.6 (Thinking) | 52.33 |
| 8 | Qwen 3.7 Max | 49.69 |
| 9 | DeepSeek V4 Flash | 48.6 |
| 10 | Gemini 3.5 Flash | 45.65 |
| 11 | GLM-5 (Thinking) | 43.32 |
| 12 | Seed 2.1 Pro | 42.7 |
| 13 | Qwen 3.6 27B | 42.08 |
| 14 | Gemini 3.1 Pro (Preview) | 41.3 |
| 15 | Qwen 3.5 397B A17B | 41.15 |
Interactive version: aibenchmarks.dev/benchmark?slug=omniabench · How the rankings work · Data refreshed daily, snapshot 2026-07-19.