CFEval: leaderboard
CFEval benchmark for evaluating code generation and problem-solving capabilities.
Metric: Score (points). Source: huggingface.co. Source: benchmarklist.com. Saturation forecast: Not forecast. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 235B A22B 2507 (Thinking) | 2134 | #260 (Qwen 3 235B A22B 2507) |
| 2 | DeepSeek R1 0528 | 2099 | #222 |
| 3 | Qwen 3 235B A22B (Thinking) | 2056 | #313 (Qwen 3 235B A22B) |
| 4 | O3 | 2043 | #124 |
| 5 | Gemini 2.5 Pro | 2001 | #152 |
| 6 | O4 Mini | 1929 | #180 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=cfeval · How It Works · Data refreshed daily, snapshot 2026-10-20.