Qwen2.5-VL-7B-Instruct
Alibaba · Qwen/Qwen2.5-VL-7B-Instruct
Weights license: Apache 2.0 · code and components may use other terms
Quality
No comparable score
We track no public, comparable benchmark that covers this model yet. No score is inferred.
Reported by the publisher
- MMMU val
- 58.6Image benchmark
- MMMU-Pro val
- 41.0Image benchmark
- DocVQA test
- 95.7Image benchmark
- InfoVQA test
- 82.6Image benchmark
- ChartQA test
- 87.3Image benchmark
- TextVQA val
- 84.9Image benchmark
- OCRBench
- 864Image benchmark
- CC_OCR
- 77.8Image benchmark
Every reported result
- MMStar
- 63.9Image benchmark
- MMBench-V1.1-En test
- 82.6Image benchmark
- MMT-Bench test
- 63.6Image benchmark
- MMVet GPT-4-Turbo
- 67.1Image benchmark
- HallBench avg
- 52.9Image benchmark
- MathVista testmini
- 68.2Image benchmark
- MathVision
- 25.07Image benchmark
- MVBench
- 69.6Video Benchmarks
- PerceptionTest test
- 70.5Video Benchmarks
- LVBench
- 45.3Video Benchmarks
- LongVideoBench
- 54.7Video Benchmarks
- MMBench-Video
- 1.79Video Benchmarks
- TempCompass
- 71.7Video Benchmarks
- MLVU
- 70.2Video Benchmarks
- ScreenSpot
- 84.7Agent benchmark
- ScreenSpot Pro
- 29.0Agent benchmark
- AITZ_EM
- 81.9Agent benchmark
- Android Control High_EM
- 60.1Agent benchmark
- Android Control Low_EM
- 93.7Agent benchmark
- AndroidWorld_SR
- 25.5Agent benchmark
- MobileMiniWob++_SR
- 91.4Agent benchmark
Values as printed on the card, with the publisher's own settings (card section: Evaluation › Image benchmark; Evaluation › Video Benchmarks); column “Qwen2.5-VL-7B”. Not comparable across publishers and not used in any ranking. Model card
Runs on
Fast memory holds it 90
The planning estimate fits the memory the model can use quickly: graphics card VRAM, or unified memory within the macOS GPU share.
Show all 90
Within memory, outside the fast part 26
Within the machine's memory, outside its fast part: on a graphics card the layers that do not fit run on the CPU from PC RAM (the listed RAM, or 32 GB assumed); on a Mac, above the macOS default GPU share. Not a speed figure.
Show all 26
One memory tier short 3
The same machines with the next memory size up would hold it in fast memory.
Hardware profiles grouped by the Q4 planning estimate on each machine (dedicated GPUs reserve 2 GB; unified and system memory 25%, 2 to 8 GB): the fast memory holds it, it is within memory but outside the fast part (layers offloaded to PC RAM, or above the macOS GPU share), or the machine is one memory tier short. Not a measured fit or speed.
Alibaba released Qwen2.5-VL-7B-Instruct on 2025-01-26: a vision-language model that answers about images, with 8.29B parameters on the Hugging Face count and the Apache 2.0 license. Languages the publisher declares: en.
License: commercial use allowed. At Q4 the weights load about 6.2 GB before each machine's reserve; image resolution, video frames and context add memory, and the vision encoder needs support in the runtime you pick.
Measured speed
No published measurement for this checkpoint
Tokens per second are shown only when a source publishes them with hardware, build and setup. They are never estimated.
How to run
Runtime compatibility and tokens per second have not been measured for this entry. Published metadata and a memory estimate do not prove a successful local run.
Library in source: transformers
Documented ways to run it
Copied as printed in the linked source (read 2026-10-18); not run by this site.
# It's highly recommanded to use `[decord]` feature for faster video loading.
pip install qwen-vl-utils[decord]==0.0.8from transformers import Qwen2_5_VLForConditionalGeneration, AutoTokenizer, AutoProcessor
from qwen_vl_utils import process_vision_info
# default: Load the model on the available device(s)
model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
"Qwen/Qwen2.5-VL-7B-Instruct", torch_dtype="auto", device_map="auto"
)
# We recommend enabling flash_attention_2 for better acceleration and memory saving, especially in multi-image and video scenarios.
# model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
# "Qwen/Qwen2.5-VL-7B-Instruct",
# torch_dtype=torch.bfloat16,
# attn_implementation="flash_attention_2",
# device_map="auto",
# )
# default processer
processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-7B-Instruct")
# The default range for the number of visual tokens per image in the model is 4-16384.
# You can set min_pixels and max_pixels according to your needs, such as a token range of 256-1280, to balance performance and cost.
# min_pixels = 256*28*28
# max_pixels = 1280*28*28
# processor = AutoProcessor.from_pretrained("Qwen/Qwen2.5-VL-7B-Instruct", min_pixels=min_pixels, max_pixels=max_pixels)
messages = [
{
"role": "user",
"content": [
{
"type": "image",
"image": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo.jpeg",
},
{"type": "text", "text": "Describe this image."},
],
}
]
# Preparation for inference
text = processor.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
text=[text],
images=image_inputs,
videos=video_inputs,
padding=True,
return_tensors="pt",
)
inputs = inputs.to("cuda")
# Inference: Generation of the output
generated_ids = model.generate(**inputs, max_new_tokens=128)
generated_ids_trimmed = [
out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
output_text = processor.batch_decode(
generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)
print(output_text)Technical details
What the source and the catalog store for this checkpoint. Missing fields are left out, never zero.
- Checkpoint / source
- Qwen/Qwen2.5-VL-7B-Instruct
- Task
- image-text-to-text
- Architecture
- Qwen2_5_VLForConditionalGeneration
- Library in source
- transformers
- Weight license
- Apache 2.0Qwen/Qwen2.5-VL-7B-Instruct model card (Hugging Face) · read 2026-10-18source
- Commercial use
- Allowed Apache License 2.0 text (SPDX) · read 2026-10-18license text
- Languages
- enModel card metadata · read 2026-10-25source
- Released
- Qwen2.5-VL blog (Qwen), published January 26, 2025: open-sources 3B, 7B and 72B base and instruct models on Hugging Face and ModelScope · read 2026-10-10source
Every stored field
- Revision checked
- cc594898137f460bfe9f0759e9844b3ce807cfb5
- Architecture
- Qwen2_5_VLForConditionalGeneration
- Library
- transformers
- Hub task tag
- image-text-to-text
- Formats
- transformers, safetensors
- Access
- Open download
- Publisher kind
- Official
- Hub parameter count
- 8,292,166,656safetensors.total
- Repository created
- 2025-01-26creation is not the model's release
- Repository updated
- 2025-04-06a commit is not a release date
- Hub downloads (30 days) · likes
- 5,133,806 · 1,753this repository only · read 2026-10-25
- Catalog state
- Hub config or weights found; no inventory benchmark link
Sources
Identity and published specifications come from the checkpoint page. Benchmark scores belong to the named release and do not measure speed on your computer.
- Publisher model card
- License: Qwen/Qwen2.5-VL-7B-Instruct model card (Hugging Face)
- Release date: Qwen2.5-VL blog (Qwen), published January 26, 2025: open-sources 3B, 7B and 72B base and instruct models on Hugging Face and ModelScope
Curated facts checked: 2026-10-18. The license and release date cite the publisher's sources; parameters, files and task come from the Hugging Face record, and memory is a labelled estimate from its parameter count. A value no source states stays empty; a repository commit is not a model release date.






























































