List your company

ResearchFormerly WizWand

User Interface Reasoning on UXBench

79Bubble Occurrence TUI-UX

Evaluation results

Method	Links
UI-UX2026.06		79	88	79	77	94	72	71	76	79.63
Claude-4.5-Sonnet2026.06		65	55	77	53	66	70	66	72	65.5
Claude-3.7-Sonnet2026.06		64	66	78	56	53	61	65	76	64.88
Claude-4-Sonnet2026.06		64	52	77	46	52	68	66	74	62.38
MimoVL-05282026.06		58	56	74	41	52	54	40	60	54.38
Qwen3-VL-Thinking2026.06		52	61	82	47	56	54	64	70	58.54
Qwen3-VL-Thinking2026.06		47	49	72	39	50	52	63	65	52.54
InternVL-3.52026.06		42	34	43	26	37	61	72	76	48.88
Qwen3-VL2026.06		30	57	79	40	52	62	64	70	56
InternVL-32026.06		30	34	61	0	34	52	11	8	28.75
MiniCPM-V-4.52026.06		30	39	56	33	55	53	65	76	50.86
Qwen2.5-VL2026.06		9	51	70	55	56	69	68	71	54.82
Llava3-Next2026.06		0	10	22	22	26	19	0	0	12
GLM-4.1-Thinking2026.06		0	60	62	34	32	32	4	1	28.13
InternVL-3.52026.06		0	39	65	38	50	59	68	72	48.88