Smolagents - MATH (Tool Calling): leaderboard
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 6 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4o | 58 |
| 2 | Claude 3.5 Sonnet | 54 |
| 3 | DeepSeek R1 Distill Qwen 32B | 44 |
| 4 | Qwen 2.5 Coder 32B Instruct | 44 |
| 5 | Llama 3.1 70B Instruct | 16 |
| 6 | Llama 3.1 8B Instruct | 12 |
Interactive version: theaggregate.ai/benchmark?slug=smolagents-math-tool-calling · How It Works · Data refreshed daily, snapshot 2026-10-09.