Qwen 2.5 VL — local AI model by Alibaba Cloud
Written by Jakub Rusinowski · Last updated
Alibaba's powerful vision-language models that understand images, documents, charts, and videos alongside text. Qwen 2.5 VL 72B rivals GPT-4V on multiple visual benchmarks, while the 7B version delivers strong OCR and document parsing in under 6GB VRAM.
Variants
The smallest Qwen 2.5 VL variant needs about 6 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache.
| Model | VRAM at Q4 | VRAM | Context | Run it |
|---|---|---|---|---|
| Qwen 2.5 VL 7B Instruct → 7B | ~5.8 GB | 128,000 | ollama run qwen2.5vl:7b | |
| Qwen 2.5 VL 72B Instruct → 72B | ~45.1 GB | 128,000 | ollama run qwen2.5vl:72b |
Memory is quantized weights plus overhead at Q4_K_M, from the same engine as the GPU & VRAM checker.
How to run Qwen 2.5 VL locally
Install Ollama, then pull the tag.
ollama run qwen2.5vl:7bPick a size above for its own VRAM figure, speed estimate and install command.
Licence
Commercial use permitted. No usage restrictions beyond attribution.
Applies to: Qwen 2.5 VL 7B Instruct, Qwen 2.5 VL 72B InstructRecommended GPU
The cheapest catalogued GPU that runs Qwen 2.5 VL locally (min 6 GB VRAM) is the Intel Arc B570 (10 GB).
Can I run Qwen 2.5 VL on my GPU?
- Qwen 2.5 VL on AMD Radeon RX 9060 XT 8GB
- Qwen 2.5 VL on AMD Ryzen AI Max+ 395
- Qwen 2.5 VL on Apple M1 Max
- Qwen 2.5 VL on Apple M2 Max
- Qwen 2.5 VL on Apple M3 Max
- Qwen 2.5 VL on Apple M5 Pro
- Qwen 2.5 VL on Intel Arc B570
- Qwen 2.5 VL on Intel Arc B580
- Qwen 2.5 VL on NVIDIA A100 80GB (PCIe)
- Qwen 2.5 VL on NVIDIA GeForce RTX 3060 (12GB)
- Qwen 2.5 VL on NVIDIA GeForce RTX 3070
- Qwen 2.5 VL on NVIDIA GeForce RTX 3070 Ti
- Qwen 2.5 VL on NVIDIA GeForce RTX 3080 (10GB)
- Qwen 2.5 VL on NVIDIA GeForce RTX 4060
- Qwen 2.5 VL on NVIDIA GeForce RTX 4070
- Qwen 2.5 VL on NVIDIA GeForce RTX 4070 Super
- Qwen 2.5 VL on NVIDIA GeForce RTX 4070 Ti
- Qwen 2.5 VL on NVIDIA GeForce RTX 5060
- Qwen 2.5 VL on NVIDIA GeForce RTX 5060 Ti 8GB
- Qwen 2.5 VL on NVIDIA GeForce RTX 5070
- Qwen 2.5 VL on NVIDIA H100 80GB (PCIe)
- Qwen 2.5 VL on NVIDIA L40S
- Qwen 2.5 VL on NVIDIA RTX 6000 Ada Generation
- Qwen 2.5 VL on NVIDIA RTX PRO 6000 Blackwell
Qwen 2.5 VL — frequently asked questions
How much VRAM does Qwen 2.5 VL need?
Qwen 2.5 VL needs about 6 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: Qwen 2.5 VL 7B Instruct (6 GB, Q4_K_M); Qwen 2.5 VL 72B Instruct (45 GB, Q4_K_M). On Apple Silicon, unified memory counts toward this requirement.
Can I run Qwen 2.5 VL on an RTX 4090 (24 GB)?
Yes — Qwen 2.5 VL runs on an RTX 4090 (24 GB) and other 24 GB cards such as the RTX 3090. Smaller variants also fit comfortably on 8–16 GB GPUs at Q4_K_M.
What quantization should I use for Qwen 2.5 VL?
Q4_K_M is the best balance of quality and VRAM for Qwen 2.5 VL in most cases. Choose Q8_0 for near-lossless quality if you have spare VRAM, or smaller quants (Q3/Q2) only when memory is tight.
How do I run Qwen 2.5 VL with Ollama?
Install Ollama, then run: ollama run qwen2.5vl:7b. This downloads Qwen 2.5 VL and starts a local, OpenAI-compatible endpoint — no internet connection is needed after the initial download.