Smolagents - SimpleQA (Code Agent): leaderboard
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 15 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 2.5 72B Instruct | 88 |
| 2 | GPT-4.5 (Preview) | 88 |
| 3 | Claude 3.7 Sonnet (20250219) | 86 |
| 4 | O1 | 84 |
| 5 | Claude 3.5 Sonnet | 82 |
| 6 | GPT-4o | 80 |
| 7 | O3 Mini | 80 |
| 8 | Llama 3.3 70B Instruct | 78 |
| 9 | DeepSeek R1 | 74 |
| 10 | QwQ-32B | 74 |
| 11 | Qwen 2.5 Coder 32B Instruct | 74 |
| 12 | Llama 3.1 70B Instruct | 64 |
| 13 | DeepSeek R1 Distill Qwen 32B | 40 |
| 14 | Llama 3.2 3B Instruct | 20 |
| 15 | Llama 3.1 8B Instruct | 2 |
Interactive version: theaggregate.ai/benchmark?slug=smolagents-simpleqa-code-agent · How It Works · Data refreshed daily, snapshot 2026-10-09.