Vision-language · Alibaba Qwen

Qwen2.5 VL 7B

Hugging Face: Qwen/Qwen2.5-VL-7B-Instruct · checked 2026-10-13

Image, document and video understanding at 7B. A common base for OCR tunes.

Datasheet

Maker	Alibaba Qwen
Parameters	8.29 BVerifiedsafetensors.total from the HF API Source
Licence	Apache-2.0Verified
Access	Public download
Task (HF pipeline)	image-text-to-text
First published on HF	2025-01-26
Last updated on HF	2025-04-06
Weights in memory	17 GB at 16-bit · 5.0 GB at ~4.85 bitsEstimateParameters × bits ÷ 8. Excludes context cache and runtime overhead; not every model has a 4-bit build.

Benchmark

58.6%

MMMU (val) · #1 of 3 scored in this category

MMMU val, model card table. Source

See the full vision-language ranking

Used in

Smarter camera alerts
A patient maths tutor

The description summarises the model card at https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct. Parameters and licence are read from the Hugging Face API; the weights-in-memory line is our arithmetic, not a measurement.