FactArena - Justification Informativeness: leaderboard
Metric: Elo rating from pairwise judged comparisons. Source: arxiv.org. Saturation forecast: Estimated already saturated. 16 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O3 (2025-04-16) | 1331.77 |
| 2 | DeepSeek R1 | 1160.02 |
| 3 | O4 Mini (2025-04-16) | 1101.49 |
| 4 | Gemini 2.5 Pro (Preview 06-05) | 1078.02 |
| 5 | GPT-4.5 (Preview) | 1039.95 |
| 6 | GPT-4.1 | 1032.77 |
| 7 | Grok 3 | 1026.11 |
| 8 | Grok 3 Mini Beta | 966.4 |
| 9 | Claude Opus 4 (20250514) | 961.3 |
| 10 | Gemini 2.5 Flash | 945.27 |
| 11 | Claude Sonnet 4 (20250514) | 930.52 |
| 12 | Llama 4 Maverick Instruct | 926.36 |
| 13 | DeepSeek V3 | 923.45 |
| 14 | Qwen 3 235B A22B | 882.68 |
| 15 | GPT-4o | 872.55 |
Interactive version: theaggregate.ai/benchmark?slug=factarena-justification-informativeness · How It Works · Data refreshed daily, snapshot 2026-09-25.