id: qwen2_5-vl-7b-instruct
name: Qwen2.5-VL 7B Instruct
type: vision-language-model
task:
  - video
  - image-understanding
  - document-understanding
  - visual-question-answering
language:
  - ru
  - en
  - multilingual
source: huggingface
upstream_id: Qwen/Qwen2.5-VL-7B-Instruct
upstream_url: https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct
license: apache-2.0
status: candidate
created_at: 2026-06-19
storage_path: /models/video/qwen2.5-vl-7b-instruct
format: safetensors
quantization: none
context_length: null
vram_required_gb: null
parameters: 7B
base_model: Qwen/Qwen2.5-VL
adapter_for: null
datasets: []
eval_suites: []
deployment:
  target: docker-gpu.cin.su
  runtime: transformers
  served_model_name: qwen2.5-vl-7b-instruct
notes: "Video/image understanding candidate for the video plugin. Supports long-video understanding according to the model card."