Equational Theories - Hard2 (Default reasoning): leaderboard
Metric: Strict F1 (%). Source: huggingface.co. Saturation forecast: Around January 2027. 25 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 | 92.13 |
| 2 | Gemini 3.1 Pro (Preview) | 90.42 |
| 3 | Grok 4.1 Fast | 76.43 |
| 4 | Kimi K2.5 | 74.53 |
| 5 | Gemini 3 Flash (Preview) | 73.14 |
| 6 | Qwen 3.5 397B A17B | 72.94 |
| 7 | GLM-5 | 72.21 |
| 8 | GPT-5 Mini | 71.99 |
| 9 | Qwen 3.5 122B A10B | 71.78 |
| 10 | Qwen 3.5 27B | 71.64 |
| 11 | Claude Opus 4.6 | 70.86 |
| 12 | GPT-OSS-120B | 68.61 |
| 13 | Claude Haiku 4.5 | 65.55 |
| 14 | Seed 2.0 Lite | 62.08 |
| 15 | Llama 3.1 8B Instruct | 58.9 |
Interactive version: theaggregate.ai/benchmark?slug=equational-theories-hard2-default-reasoning · How It Works · Data refreshed daily, snapshot 2026-10-09.