FactArena - Justification Readability: 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) | 1268.83 |
| 2 | DeepSeek R1 | 1117.53 |
| 3 | O4 Mini (2025-04-16) | 1097.51 |
| 4 | Gemini 2.5 Pro (Preview 06-05) | 1077.44 |
| 5 | GPT-4.5 (Preview) | 1041.03 |
| 6 | GPT-4.1 | 1029.11 |
| 7 | Grok 3 | 1019.52 |
| 8 | Gemini 2.5 Flash | 980.5 |
| 9 | Grok 3 Mini Beta | 974.06 |
| 10 | DeepSeek V3 | 951.61 |
| 11 | Claude Opus 4 (20250514) | 933.39 |
| 12 | Qwen 3 235B A22B | 922.65 |
| 13 | GPT-4o | 913.03 |
| 14 | Llama 4 Maverick Instruct | 909.13 |
| 15 | Claude Sonnet 4 (20250514) | 892.76 |
Interactive version: theaggregate.ai/benchmark?slug=factarena-justification-readability · How It Works · Data refreshed daily, snapshot 2026-09-25.