Image Editing on Arena Analysis March 26, 2026 (test)
1,270Arena ELOGPT Image 1.5 (26-03-04)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GPT Image 1.5 (26-03-04)Samples=1,081, Evaluation Protocol=Human Annotation2026.03 | 1,270 | 1 | |
| Nano Banana Pro (26-03-04)Samples=1,156, Evaluation Protocol=Human Annotation2026.03 | 1,251 | 2 | |
| Seedream 4.5 (26-03-11)Samples=1,190, Evaluation Protocol=Human Annotation2026.03 | 1,196 | 3 | |
| FLUX.2 [klein] 9BSamples=1,200, Evaluation Protocol=Human Annotation2026.03 | 1,166 | 4 | |
| Qwen-Image-Edit-2511Samples=1,200, Evaluation Protocol=Human Annotation2026.03 | 1,164 | 5 | |
| FLUX.2 [dev] TurboSamples=1,200, Evaluation Protocol=Human Annotation2026.03 | 1,153 | 6 | |
| Qwen-Image-Edit-2509Samples=1,200, Evaluation Protocol=Human Annotation2026.03 | 1,142 | 7 | |
| FLUX.2 [dev]Samples=1,200, Evaluation Protocol=Human Annotation2026.03 | 1,137 | 8 | |
| LongCat-Image-EditSamples=1,200, Evaluation Protocol=Human Annotation2026.03 | 1,111 | 9 | |
| FLUX.2 [klein] 4BSamples=1,200, Evaluation Protocol=Human Annotation2026.03 | 1,107 | 10 | |
| Step1X-Edit-v1p2Samples=1,200, Evaluation Protocol=Human Annotation2026.03 | 1,093 | 11 | |
| Qwen-Image-EditSamples=1,200, Evaluation Protocol=Human Annotation2026.03 | 1,088 | 12 | |
| FLUX.1 Kontext [dev]Samples=1,200, Evaluation Protocol=Human Annotation2026.03 | 1,017 | 13 | |
| GLM-ImageSamples=1,200, Evaluation Protocol=Human Annotation2026.03 | 930 | 14 | |
| OmniGen V2Samples=1,200, Evaluation Protocol=Human Annotation2026.03 | 919 | 15 | |
| BagelSamples=1,200, Evaluation Protocol=Human Annotation2026.03 | 915 | 16 |