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

Fact-based Visual Question Answering on FVQA (test)

72.61Score

Gemini-3-Pro

Updated 19d ago

Evaluation Results

Method	Links
Gemini-3-Pro 2026.07		72.61	-	-	-
GPT-5.2 2026.07		68.78	-	-	-
GPT-4o 2026.07		66.34	-	-	-
Gemini-3-Flash 2026.07		64.89	-	-	-
VideoSearcher-8B 2026.07		64.1	-	-	-
DeepEyesV2 2026.07		60.6	-	-	-
Gemini-3-Pro 2026.07		59.22	-	-	-
MMSearch-R1 2026.07		58.4	-	-	-
Webwatcher-7B 2026.07		58.17	-	-	-
Gemini-3-Flash 2026.07		56.5	-	-	-
VideoSearcher-4B 2026.07		55.17	-	-	-
Qwen3-VL-32B-Instruct 2026.07		54.28	-	-	-
Qwen3-VL-8B-Instruct 2026.07		53.61	-	-	-
Qwen2.5-VL-32B-Instruct 2026.07		52.22	-	-	-
GPT-5.2 2026.07		50.94	-	-	-
GPT-4o 2026.07		48	-	-	-
Visual-ARFT 2026.07		41.72	-	-	-
Qwen2.5-VL-7B-Instruct 2026.07		36	-	-	-
Qwen3-VL-32B-Instruct 2026.07		32.17	-	-	-
Qwen3-VL-4B-Instruct 2026.07		31.72	-	-	-
Qwen2.5-VL-32B-Instruct 2026.07		30.5	-	-	-
Qwen2.5-VL-7B-Instruct 2026.07		26.28	-	-	-
Qwen3-VL-4B-Instruct 2026.07		25.33	-	-	-
Qwen3-VL-8B-Instruct 2026.07		24.22	-	-	-
SVM-Question 2016.06		-	17.06	29.43	44.73
SVM-Image 2016.06		-	24.73	40.95	56.33
SVM-Question+Image 2016.06		-	25.3	41.37	56.78
LSTM-Question 2016.06		-	15.82	26.45	40.99
LSTM-Image 2016.06		-	26.78	44	62.86
LSTM-Question+Image 2016.06		-	29.08	44.36	61.71
LSTM-Question+Image+Pre-VQA 2016.06		-	31.96	48.55	64.73
Hie-Question+Image 2016.06		-	39.75	56.48	69.78
Hie-Question+Image+Pre-VQA 2016.06		-	48.93	64.75	76.73
FVQA model 2016.06		-	65.51	72.37	73.55
FVQA model 2016.06		-	54.79	61.41	62.22
FVQA model 2016.06		-	59.67	66.89	67.77
Human 2016.06		-	82.47	-	-