Share your thoughts, 1 month free Claude Pro on usSee more

Reasoning-aware Image Generation on RiseBench 1.0 (test)

0.77Instruction Reasoning

Gemini-3-pro-image-preview

Updated 1mo ago

Evaluation Results

Method	Links
Gemini-3-pro-image-preview 2026.01		0.77	0.855	0.944	0.412	0.611	0.48	0.376	0.472
GPT-Image-1 2026.01		0.628	0.802	0.949	0.341	0.322	0.37	0.106	0.289
Qwen-Image-Edit + Unified Thinker 2026.01		0.619	0.762	0.905	0.329	0.3	0.41	0.094	0.289
Gemini-2.5-Flash-Image 2026.01		0.612	0.86	0.913	0.259	0.478	0.37	0.188	0.328
Seedream-4.0 2026.01		0.589	0.674	0.912	0.129	0.122	0.11	0.071	0.108
BAGEL + Unified Thinker 2026.01		0.587	0.757	0.809	0.152	0.177	0.2	0.082	0.155
Qwen-Image-Edit + Unified Thinker 2026.01		0.586	0.759	0.901	0.247	0.222	0.38	0.094	0.242
GPT-Image-1-mini 2026.01		0.541	0.715	0.937	0.247	0.289	0.33	0.094	0.244
BAGEL + Unified Thinker 2026.01		0.533	0.736	0.781	0.141	0.177	0.18	0.035	0.136
Gemini-2.0-Flash-pre 2026.01		0.499	0.684	0.849	0.106	0.133	0.11	0.023	0.094
Gemini-2.0-Flash-exp 2026.01		0.489	0.682	0.827	0.082	0.155	0.23	0.047	0.133
BAGEL 2026.01		0.459	0.738	0.801	0.059	0.178	0.21	0.012	0.119
Qwen-Image-Edit 2026.01		0.372	0.664	0.869	0.047	0.1	0.17	0.024	0.089
BAGEL 2026.01		0.365	0.535	0.73	0.024	0.056	0.14	0.012	0.061
Ovis-U1 2026.01		0.339	0.527	0.729	0.012	0.033	0.04	0.024	0.028
Step1X-Edit 2026.01		0.303	0.126	0.749	0	0.022	0.02	0.035	0.019
FLUX.1-Kontext-Dev 2026.01		0.26	0.716	0.852	0.023	0.055	0.13	0.012	0.058
OmniGen 2026.01		0.251	0.415	0.735	0.012	0.01	0	0.012	0.008
EMU2 2026.01		0.226	0.382	0.783	0.012	0.011	0	0	0.005