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

Text-centric Reasoning on VideoThinkBench mini (test)

89Average Score

Gemini 2.5 Pro

Updated 1mo ago

Evaluation Results

Method	Links
Gemini 2.5 Pro 2025.11		89	100	100	100	90	100	83.3	93.3	80	80	85	65	95	85
GPT-5 high 2025.11		86.6	100	100	100	100	100	83.3	93.3	80	83.9	75	55	85	70
Qwen3-VL-235B-A22B 2025.11		77.6	100	100	80	50	84.6	58.3	100	56	80	70	65	90	75
Claude Sonnet 4.5 2025.11		77.2	100	100	60	40	100	83.3	100	60	80	80	45	80	75
Qwen3-VL-Plus 2025.11		75.8	100	95	100	70	76.9	66.7	80	64	57.1	65	65	80	65
Qwen3-VL-32B 2025.11		72.5	100	95	80	50	76.9	66.7	93.3	40	65.7	75	45	90	65
Sora-2 2025.11		67.6	100	90	50	40	76.9	66.7	73.3	56	45.7	75	45	90	70
Nano Banana Pro 2025.11		66	56.7	65	80	80	69.2	75	80	44	62.9	75	45	75	50
Sora-2 2025.11		57.1	76.7	65	40	30	69.2	66.7	73.3	52	54.3	70	45	60	40
Seedream 4.5 2025.11		55.7	100	80	20	10	69.2	75	60	36	48.6	55	60	55	55
Veo 3.1 2025.11		48.3	80	70	50	20	61.5	16.7	60	52	42.9	50	45	35	45
Veo 3.1 2025.11		44.5	93.3	80	50	20	61.5	41.7	80	40	51.4	25	5	20	10
GPT Image 1.5 2025.11		41.4	90	40	0	0	69.2	25	46.7	40	22.9	50	40	65	50
MiniMax Hailuo 2.3 2025.11		38.4	76.6	40	10	20	61.5	33.3	86.6	16	65.7	30	10	30	20
MOVA-720p 2025.11		12.5	30	35	20	0	15.4	0	6.7	8	2.9	0	25	10	10
Qwen-Image-Edit-2511 2025.11		10.5	0	0	0	0	15.4	16.7	6.7	4	8.6	15	15	30	25
MOVA-360p 2025.11		10.4	20	10	0	20	15.4	0	0	20	0	5	30	5	10
BAGEL 2025.11		8.9	6.7	0	0	0	0	0	6.7	4	2.9	25	40	10	20
MOVA-720p 2025.11		7.2	0	0	0	0	30.8	16.7	0	8	2.9	0	10	20	5
MOVA-360p 2025.11		5.8	0	0	0	0	30.8	33.3	0	0	5.7	0	0	5	0
Seedance 1.0 Pro 2025.11		0	0	0	0	0	0	0	0	0	0	0	0	0	0
Wan2.2-TI2V-5B 2025.11		0	0	0	0	0	0	0	0	0	0	0	0	0	0