LiveFact: leaderboard
Metric: Average (%) of the twelve LiveFact November 2025 scores (accuracy and macro-F1 in Classification and Inference modes at evidence offsets of -3, 0 and +3 days), 4,392 claims from 737 news events, temperature 0; higher is better. Source: arxiv.org. Saturation forecast: Around April 2027. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3 235B A22B 2507 Instruct | 72.4 |
| 2 | GPT-OSS-120B | 72.13 |
| 3 | GPT-5.1 | 72.02 |
| 4 | GPT-5.2 | 71.52 |
| 5 | Qwen 3 30B A3B 2507 Instruct | 69.46 |
| 6 | GPT-4o (2024-08-06) | 67.11 |
| 7 | GPT-4o Mini (2024-07-18) | 66.12 |
| 8 | Qwen 3 32B | 66.03 |
| 9 | Qwen 3 8B | 63.62 |
| 10 | DeepSeek V3.1 | 61.48 |
| 11 | GPT-OSS-20B | 57.05 |
| 12 | Llama 3.3 70B Instruct | 55.85 |
| 13 | Qwen 3 4B 2507 Instruct | 55.46 |
| 14 | Kimi K2 (Thinking) | 54.6 |
| 15 | Llama 3.1 8B Instruct | 49.96 |
Interactive version: theaggregate.ai/benchmark?slug=livefact · How It Works · Data refreshed daily, snapshot 2026-10-07.