FormationEval — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 72 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 99.8 |
| 2 | GLM-4.7 | 98.6 |
| 3 | Gemini 3 Flash (Preview) | 98.2 |
| 4 | Gemini 2.5 Pro | 97.8 |
| 5 | Grok 4.1 Fast | 97.6 |
| 6 | Kimi K2 (Thinking) | 97.2 |
| 7 | Claude Opus 4.5 | 97 |
| 8 | GPT-5 Mini (Medium) | 96.4 |
| 9 | DeepSeek R1 | 96.2 |
| 10 | Grok 4 Fast | 96 |
| 11 | GPT-5 Mini (High) | 95.6 |
| 12 | O4 Mini (High) | 95.2 |
| 13 | GPT-5 Mini (Low) | 95.2 |
| 14 | Gemini 2.5 Flash | 95 |
| 15 | Grok 3 Mini | 95 |
Interactive version: theaggregate.ai/benchmark?slug=formationeval · How the rankings work · Data refreshed daily, snapshot 2026-07-22.