ArgBench (Chain-of-Thought) - Argument Reasoning: leaderboard
Metric: Macro F1 (0-1, times 100) averaged over the argument reasoning (fallacies, warrants and similar) tasks, chain-of-thought prompting; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4.1 | 46.5 |
| 2 | Mistral-Small-Instruct-2409 | 43 |
| 3 | Llama 3.3 70B Instruct | 39.6 |
| 4 | Llama 3.1 8B Instruct | 38.9 |
| 5 | Mixtral 8x7B Instruct (v0.1) | 38.9 |
| 6 | Qwen 3 32B | 35.8 |
| 7 | Mistral 7B Instruct (v0.3) | 35.1 |
| 8 | Qwen 3 4B | 27.8 |
| 9 | DeepSeek R1 Distill Qwen 32B | 27.4 |
| 10 | DeepSeek-R1-Distill-Qwen-7B | 27 |
Interactive version: theaggregate.ai/benchmark?slug=argbench-chain-of-thought-argument-reasoning · How It Works · Data refreshed daily, snapshot 2026-10-07.