HELM Lite - MATH Level 1 - Prealgebra: leaderboard
Metric: Equivalent (CoT) (%). Source: crfm.stanford.edu. 91 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude 3.5 Haiku (20241022) | 98.84 |
| 2 | Claude 3.5 Sonnet (20241022) | 97.67 |
| 3 | GPT-4o (2024-05-13) | 97.67 |
| 4 | Qwen 2.5 72B Instruct | 96.51 |
| 5 | Gemini 2.0 Flash (Preview) | 96.51 |
| 6 | Gemini 1.5 Flash (002) | 95.35 |
| 7 | Claude 3.5 Sonnet (20240620) | 95.35 |
| 8 | GPT-4o Mini (2024-07-18) | 95.35 |
| 9 | GPT-4 Turbo | 95.35 |
| 10 | Gemini 1.5 Pro (001) | 95.35 |
| 11 | Llama 3.3 70B Instruct | 94.19 |
| 12 | Gemini 1.5 Pro (002) | 94.19 |
| 13 | DeepSeek V3 | 94.19 |
| 14 | GPT-4o (2024-08-06) | 93.02 |
| 15 | Gemma 2 27B (IT) | 93.02 |
Interactive version: theaggregate.ai/benchmark?slug=helm-lite-math-level-1-prealgebra · How It Works · Data refreshed daily, snapshot 2026-09-19.