Text-to-Image Generation on WISE (test) benchmark leaderboard · SOTA2 Research

Text-to-Image Generation on WISE (test)

80Overall ScoreGPT-4o

Evaluation Results

Method	Links
GPT-4o2025.05		80	81	71	89	83	79	74
GPT-4o2026.02		80	81	71	89	83	79	74
GPT-4o2026.05		80	81	71	89	83	79	74
Seedream 4.02026.02		78	78	73	85	79	84	67
UniReason2026.02		78	80	68	79	77	83	81
IRG2026.02		77	78	72	76	81	82	78
CLVR2026.05		76	76	73	83	76	79	67
UniCoT2026.02		75	76	70	76	73	81	73
CLVR2026.05		74	73	72	83	73	80	64
MindOmni2026.02		71	75	70	76	76	72	52
BAGEL2025.05		70	76	69	75	65	75	58
BAGEL2026.02		70	76	69	75	65	75	58
LongCat-Image2026.05		65	66	61	72	66	72	49
Uni-CoT2026.05		65	66	60	79	64	67	60
BLIP3-o2026.02		62	—	—	—	—	—	—
Qwen-Image2026.02		62	62	63	77	57	75	40
Qwen-Image2026.05		62	62	63	77	57	75	40
Hunyuan-Image 3.02026.02		57	58	57	70	56	63	31
Hunyuan-Image 3.02026.05		57	58	57	70	56	63	31
MetaQuery-XL2025.05		55	56	55	62	49	63	41
MetaQuery-XL2026.02		55	56	55	62	49	63	41
UniWorld-V12026.02		55	53	55	73	45	59	41
T2I-R12026.02		54	56	55	63	54	55	30
T2I-R12026.05		54	56	55	63	54	55	30
BAGEL2025.05		52	44	55	68	44	60	39
BAGEL2026.05		52	44	55	68	44	60	39
FLUX.2 9B [Klein] base2026.05		52	46	56	66	50	62	42
FLUX.1-dev2025.05		50	48	58	62	42	51	35
FLUX.1-dev2026.05		50	48	58	62	42	51	35
playground-v2.52025.05		49	49	58	55	43	48	33
PixArt-Alpha2025.05		47	45	50	48	49	56	34
OmniGen22026.02		47	42	52	64	43	50	34
SD3.5-large2025.05		46	44	50	58	44	52	31
SD3.5-Large2026.05		46	44	50	58	44	52	31
FLUX.2 4B [Klein] base2026.05		44	37	46	60	41	55	36
SDXL2025.05		43	43	48	47	44	45	27
Harmon2026.02		41	38	48	52	37	44	29
Emu32025.05		39	34	45	48	41	45	27
Show-o-5122025.05		35	28	40	48	30	46	30
Janus-Pro-7B2025.05		35	30	37	49	36	42	26
Show-o2026.02		35	28	40	48	30	46	30
Janus Pro2026.02		35	30	37	49	36	42	26
SDv1.52025.05		32	34	35	32	28	29	21
VILA-U2025.05		31	26	33	37	35	39	23
Janus2025.05		23	16	26	35	28	30	14