FMG-Bench - Perspective-Compare Harness: leaderboard
Metric: Judge-panel score (0-100; faithful disagreement surfaced). Source: github.com. 14 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 | 94.57 |
| 2 | GPT-5.4 | 94.12 |
| 3 | Kimi K2.6 | 93.82 |
| 4 | Qwen 3.6 Plus | 93.18 |
| 5 | Grok 4.20 | 92.94 |
| 6 | DeepSeek V4 Pro | 91.93 |
| 7 | GLM-5.1 | 91.34 |
| 8 | MiMo-V2.5-Pro | 90.87 |
| 9 | Gemini 3.1 Pro (Preview) | 89.75 |
| 10 | Nemotron 3 Super | 88.31 |
| 11 | MiniMax-M2.7 | 84.6 |
| 12 | Mistral Large 3 | 84.42 |
| 13 | Seed 2.0 Lite | 79.16 |
| 14 | Llama 4 Maverick | 77.09 |
Interactive version: theaggregate.ai/benchmark?slug=fmg-bench-perspective-compare-harness · How It Works · Data refreshed daily, snapshot 2026-09-19.