GenEval
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Text-to-image generation | GenEval | Overall Score96 | 467 | |
| Text-to-Image Generation | GenEval | GenEval Score95 | 277 | |
| Text-to-Image Generation | GenEval (test) | Two Obj. Acc99 | 169 | |
| Text-to-Image Generation | GenEval | Two Objects97 | 87 | |
| Text-to-Image Generation | GenEval | Overall Score88.3 | 68 | |
| Text-to-Image Generation | GenEval 1.0 (test) | Overall Score81.51 | 63 | |
| Text-to-Image Generation | GenEval++ | Color Accuracy90 | 35 | |
| Image Generation | GenEval (test) | GenEval Score91 | 35 | |
| Image Generation | GenEval overall | GenEval Overall Score90 | 30 | |
| Image Generation | GenEval | Overall Score88 | 26 | |
| Compositional Image Generation | GenEval | Overall Score0.95 | 22 | |
| Text-to-Image Generation | GenEval 1024x1024 | Latency (s)0.56 | 22 | |
| Composition Image Generation | GenEval | GenEval Score97 | 20 | |
| Text-to-Image | GenEval 11 (test) | Accuracy (Single Obj)100 | 19 | |
| Text-to-Image Generation | GenEval | DINO0.786 | 18 | |
| Text-to-Image Generation | GenEval | GenEval Score87 | 16 | |
| Text-to-Image Generation | GenEval official (val) | Object Presence (1 Obj)100 | 15 | |
| Text-to-Image Generation | GenEval | GenEval Score0.78 | 13 | |
| Text-to-Image Generation | GenEval | Single Object Score99 | 13 | |
| Text-to-Image | GenEval | Overall Score0.733 | 12 | |
| Text-to-Image Generation | GenEval | Single Object Accuracy100 | 11 | |
| Image Generation | GenEval++ | Color Accuracy90 | 10 | |
| Instruction-following generation | GenEval++ (test) | Color Accuracy90 | 9 | |
| Spatial Reasoning Generation | GenEval (test) | Mean Score90.3 | 9 | |
| Text-to-Image Synthesis | GenEval SD V1.5 | Overall Score57 | 9 |
Showing 25 of 35 rows