Qwen2.5-VL-72B.

unknown

§ 02 · Benchmarks

Every benchmark Qwen2.5-VL-72B has a recorded score for.

#	Benchmark	Area · Task	Metric	Value	Rank	Date	Source
01	HumanEval+	Computer Code · Code Generation	pass-1	87.8%	#2/12	—	source ↗
02	DocVQA	Computer Vision · Document Understanding	anls	96.4%	#5/21	—	source ↗
03	MMBench	Multimodal · Visual Question Answering	accuracy	88.6%	#7/20	—	source ↗
04	Video-MME	Multimodal · Video Understanding	accuracy	79.1%	#8/24	—	source ↗
05	TextVQA	Multimodal · Visual Question Answering	accuracy	83.5%	#9/23	—	source ↗
06	MVBench	Multimodal · Video Understanding	accuracy	70.4%	#12/20	—	source ↗
07	RealWorldQA	Multimodal · Visual Question Answering	accuracy	75.7%	#12/23	—	source ↗
08	MMStar	Multimodal · Image-Text-to-Text	accuracy	70.8%	#14/21	—	source ↗
09	MMMU	Multimodal · Image-Text-to-Text	accuracy	70.2%	#15/36	—	source ↗
10	GSM8K	Reasoning · Mathematical Reasoning	accuracy	95.3%	#23/48	—	source ↗
11	MMMU-Pro	Multimodal · Visual Question Answering	accuracy	51.1%	#24/31	—	source ↗
12	MATH	Reasoning · Mathematical Reasoning	accuracy	83.0%	#26/46	—	source ↗
13	OSWorld	Agentic AI · Web & Desktop Agents	success-rate	8.8%	#26/28	—	source ↗
14	MMLU-Pro	Reasoning · Commonsense Reasoning	accuracy	71.2%	#61/73	—	source ↗
15	GPQA Diamond	Reasoning · Multi-step Reasoning	accuracy	49.0%	#67/74	—	source ↗

Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.

§ 03 · Strengths by area