AGI-Eval Community - Algorithmic Reasoning: leaderboard
AGI-Eval community leaderboard (agi-eval.cn), the private-set LLM board built from the platform's non-public items; latest monthly snapshot. Unrelated to the AGIEval exam dataset. This board is the algorithmic (code) reasoning dimension; upstream's Chinese split is identical, so only this column is kept. Upstream accuracy 0-1, shown as 0-100; higher is better.
Metric: Accuracy (%). Source: agi-eval.cn. Saturation forecast: Around February 2027. 141 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.5 | 87.77 | #33 |
| 2 | Claude Fable 5 | 85.72 | #20 |
| 3 | Kimi K3 | 80.95 | #35 |
| 4 | Gemini 3.1 Pro (Preview) | 79.08 | #64 |
| 5 | Claude Sonnet 5 | 73.33 | #65 |
| 6 | Seed 2.1 Pro | 72.15 | #73 |
| 7 | DeepSeek V4 Pro (Max) | 71.92 | #112 (DeepSeek V4 Pro) |
| 8 | Gemini 3.5 Flash | 71.47 | #66 |
| 9 | GPT-5.4 | 71.25 | #83 |
| 10 | Gemini 3 Pro (Preview) | 70.32 | #71 |
| 11 | Gemini 3 Flash (Preview) | 69.03 | #90 |
| 12 | DeepSeek V4 Flash (Max) | 68.42 | #155 (DeepSeek V4 Flash) |
| 13 | GPT-5.1 | 67.93 | #145 |
| 14 | Seed 2.0 Pro | 67.6 | #122 |
| 15 | Claude Opus 4.8 | 67.53 | #39 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=agi-eval-community-algorithmic-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-05.