OpenEvals - GSM8K: leaderboard
Hugging Face OpenEvals official leaderboard covering GSM8K grade-school math word problems.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2 72B | 89.5 | #521 |
| 2 | DeepSeek V3 | 89.3 | #321 |
| 3 | Phi-3.5-mini-instruct | 86.2 | #1167 |
| 4 | Llama 3.1 8B Instruct | 84.5 | #1035 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=openevals-gsm8k · How It Works · Data refreshed daily, snapshot 2026-10-20.