Compositional Text-to-Image Generation on T2I-CompBench benchmark leaderboard · SOTA2 Research

Compositional Text-to-Image Generation on T2I-CompBench

0.7935Color AccuracyBiDPO

Evaluation Results

Method	Links
BiDPO2026.05		0.7935	0.6047	0.7136	0.2341	0.3229
GenTron2026.05		0.7674	0.57	0.715	0.2098	0.3202
Flux.1-dev2026.05		0.7635	0.511	0.6279	0.2802	0.308
ELLA2026.05		0.726	0.5634	0.6686	0.2214	0.3069
EVOGEN2026.05		0.7104	0.5457	0.7234	0.2176	0.3308
Dimba-G2026.05		0.6921	0.5707	0.6821	0.2105	0.3298
MARS2026.05		0.6913	0.5431	0.7123	0.1924	0.321
PixArt-α2026.05		0.6886	0.5582	0.7044	0.2082	0.3179
GORS2026.05		0.6603	0.4785	0.6287	0.1815	0.3193
Attn-Exct v22026.05		0.64	0.4517	0.5963	0.1455	0.3109
ECLIPSE2026.05		0.6119	0.5429	0.6165	0.1903	0.3139
SDXL2026.05		0.589	0.469	0.5313	0.2123	0.312
InstanceDiffusion2026.05		0.5433	0.4472	0.5293	0.2791	0.2947
Stable Diffusion 22026.05		0.5065	0.4221	0.4922	0.1342	0.3096
LMD+2026.05		0.4814	0.4865	0.5699	0.2537	0.2828
GLIGEN2026.05		0.4288	0.3998	0.3904	0.2632	0.3036