FirmCORe - Collaboration Type: leaderboard
Metric: Macro-F1 (%). Source: arxiv.org. 17 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Kimi K2.7 | 60.23 |
| 2 | Qwen 3.7 Plus | 58.1 |
| 3 | Qwen 3 4B | 57.14 |
| 4 | GLM-5.1 | 56.28 |
| 5 | Qwen 3.6 35B A3B | 54.53 |
| 6 | GLM-5.2 | 54.25 |
| 7 | Gemma 4 26B | 53.34 |
| 8 | Qwen 3.6 Max | 51.97 |
| 9 | Qwen 3.6 27B | 51.05 |
| 10 | Qwen 3 8B | 47.42 |
| 11 | Qwen3 Coder | 46.65 |
| 12 | Gemma 3 12B | 44.44 |
| 13 | Llama 3.1 8B | 42.77 |
| 14 | Llama 3.2 3B | 14.55 |
Interactive version: theaggregate.ai/benchmark?slug=firmcore-collaboration-type · How It Works · Data refreshed daily, snapshot 2026-09-20.