IDEA-Bench Arena - Full: leaderboard
IDEA-Bench text-to-image generation arena Elo from full-identity pairwise battles.
Metric: Full Arena Elo. Source: huggingface.co. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o + FLUX.1 [dev] | 1051.4 | |
| 2 | GPT-4o + PixArt-Sigma | 1014.7 | |
| 3 | GPT-4o + Stable Diffusion 3 Medium | 1012.9 | |
| 4 | GPT-4o + DALLE-3 | 1006.4 | |
| 5 | ChatDiT | 1005.3 | |
| 6 | GPT-4o + Emu2 | 890.5 | |
| 7 | GPT-4o + OmniGen | 868.4 |
Interactive version: theaggregate.ai/benchmark?slug=idea-bench-arena-full · How It Works · Data refreshed daily, snapshot 2026-10-20.