Flame Machine Learning - Create: leaderboard
Metric: Accuracy (%; the paper's Bloom-level weighted accuracy on the Create problems of the 970-problem textbook-grounded Machine Learning multiple-choice set, zero-shot, temperature 0). Source: arxiv.org. Saturation forecast: Estimated already saturated. 12 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 99 |
| 2 | GPT-5.4 | 91 |
| 3 | Claude Opus 4.6 | 88 |
| 4 | Gemini 2.5 Flash Lite | 83 |
| 5 | GPT-4.1 Mini | 82 |
| 6 | Qwen 3 32B | 75 |
| 7 | DeepSeek R1 Distill Qwen 32B | 66 |
| 8 | Claude Haiku 4.5 | 64 |
| 9 | DeepSeek R1 Distill Qwen 14B | 55 |
| 10 | Qwen 3 8B | 47 |
| 11 | Gemma 3 27B (IT) | 38 |
| 12 | Gemma 3 12B (IT) | 37 |
Interactive version: theaggregate.ai/benchmark?slug=flame-machine-learning-create · How It Works · Data refreshed daily, snapshot 2026-09-26.