AGI-Eval Community - Algorithmic Reasoning: leaderboard
Metric: Accuracy (%). Source: agi-eval.cn. 141 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 | 87.77 |
| 2 | Claude Fable 5 | 85.72 |
| 3 | Kimi K3 | 80.95 |
| 4 | Gemini 3.1 Pro (Preview) | 79.08 |
| 5 | Claude Sonnet 5 | 73.33 |
| 6 | Seed 2.1 Pro | 72.15 |
| 7 | DeepSeek V4 Pro (Max) | 71.92 |
| 8 | Gemini 3.5 Flash | 71.47 |
| 9 | GPT-5.4 | 71.25 |
| 10 | Gemini 3 Pro (Preview) | 70.32 |
| 11 | Gemini 3 Flash (Preview) | 69.03 |
| 12 | DeepSeek V4 Flash (Max) | 68.42 |
| 13 | GPT-5.1 | 67.93 |
| 14 | Seed 2.0 Pro | 67.6 |
| 15 | Claude Opus 4.8 | 67.53 |
Interactive version: theaggregate.ai/benchmark?slug=agi-eval-community-algorithmic-reasoning · How It Works · Data refreshed daily, snapshot 2026-09-19.