Text-to-Image Generation on MSCOCO 2017
23.9Pick ScoreSDXL DIDRlonger
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| SDXL DIDRlongerSteps=1, Arch.=UNet, Params=2.6B2026.05 | 23.9 | 1.1 | 33.89 | 5.83 | 32.97 | 74.99 | 0.566 | 20.6 | |
| SDXL DIDRSteps=1, Arch.=UNet, Params=2.6B2026.05 | 23.5 | 1.04 | 33.77 | 5.82 | 33.03 | 75.06 | 0.579 | 18.8 | |
| FLUX-devSteps=50, Arch.=DiT, Params=12B2026.05 | 23.2 | 1.05 | 30.65 | 5.54 | 32.88 | 83.86 | 0.672 | 24.8 | |
| SDXL Diff-Instruct*Steps=1, Arch.=UNet, Params=2.6B2026.05 | 23.1 | 1.01 | 33.29 | 5.68 | 33.09 | 74.97 | 0.555 | 18.9 | |
| SD3.5-largeSteps=28, Arch.=DiT, Params=8B2026.05 | 23 | 1.01 | 30.06 | 5.42 | 33.85 | 84.88 | 0.717 | 17.7 | |
| Z-Image-TurboSteps=8, Arch.=DiT, Params=6B2026.05 | 23 | 1.01 | 31.86 | 5.39 | 33.08 | 85 | 0.75 | 25.2 | |
| SDXLSteps=50, Arch.=UNet, Params=2.6B2026.05 | 22.8 | 0.82 | 28.98 | 5.45 | 33.86 | 74.42 | 0.551 | 15.9 | |
| SDXL-DPOSteps=50, Arch.=UNet, Params=2.6B2026.05 | 22.8 | 0.92 | 30.45 | 5.61 | 33.97 | 74.24 | 0.549 | 21.5 | |
| Z-Image DIDRSteps=1, Arch.=DiT, Params=6B2026.05 | 22.6 | 1.08 | 31.4 | 5.46 | 33.35 | 86.63 | 0.692 | 22.1 | |
| Z-ImageSteps=50, Arch.=DiT, Params=6B2026.05 | 22.5 | 0.99 | 30.57 | 5.43 | 33.15 | 86.3 | 0.84 | 14.3 | |
| SDXL Diff-InstructSteps=1, Arch.=UNet, Params=2.6B2026.05 | 22.4 | 0.92 | 31.51 | 5.53 | 33.35 | 74.98 | 0.56 | 19.3 | |
| SDXL Diff-Instruct++Steps=1, Arch.=UNet, Params=2.6B2026.05 | 22.4 | 0.92 | 31.44 | 5.58 | 33.37 | 75.01 | 0.56 | 19 | |
| SDXL DMD2Steps=1, Arch.=UNet, Params=2.6B2026.05 | 22.3 | 0.83 | 29.99 | 5.47 | 33.31 | 74.01 | 0.557 | 13.7 | |
| Z-Image-Turbo†Steps=1, Arch.=DiT, Params=6B2026.05 | 21 | 0.39 | 24.54 | 4.68 | 33.28 | 77.05 | 0.572 | 36.7 |