What it takes to serve it
GPU memory to run it for inference, with the smallest GPU setup that holds it and the lowest price on Arrakis right now.
| Precision | Memory and the smallest setup |
|---|---|
| 16-bit | 80 GB: one RTX PRO 6000, from $1.90 per GPU-hour |
| 8-bit | 40 GB: one L40S, RTX 6000 Ada or RTX A6000, L40S from $0.44 per GPU-hour |
| 4-bit | 20 GB: one RTX 4090 or RTX 3090, RTX 3090 from $0.19 per GPU-hour |
What it takes to fine-tune it
QLoRA trains adapters on a 4-bit frozen model, LoRA on a 16-bit one, and full fine-tuning updates every weight. Figures are before activations, which grow with batch size and sequence length.
| Method | Memory and the smallest setup |
|---|---|
| QLoRA | 23 GB: one RTX 4090 or RTX 3090, RTX 3090 from $0.19 per GPU-hour |
| LoRA | 80 GB: one RTX PRO 6000, from $1.90 per GPU-hour |
| Full fine-tuning | 535 GB: 4 × H200, from $16.12 per hour for all 4 |
About the model
From its card and files on Hugging Face. Memory figures here follow the GPU memory guide.
- Parameters
- 33B
- Context window (tokens)
- 128,000
- Published in
- BF16
- License
- apache-2.0
- Released
- 2025-03-21
- Weights
- Qwen/Qwen2.5-VL-32B-Instruct
Deploy it on Arrakis
Open Models in the console, choose vLLM, SGLang or llama.cpp, and rent a GPU it fits on. You get an OpenAI-compatible endpoint with its own API key, and root SSH to the machine. Or run it yourself on any machine with enough GPU memory:
vLLM
vllm serve Qwen/Qwen2.5-VL-32B-Instruct --tensor-parallel-size 1Questions
Live prices for every GPU are on the GPU prices page.
How much GPU memory does Qwen2.5-VL-32B-Instruct need?
To serve it, about 80 GB in 16-bit, 40 GB in 8-bit, 20 GB in 4-bit, including about 20% for the KV cache and activations. To fine-tune it, about 23 GB with QLoRA, 80 GB with LoRA and 535 GB for full fine-tuning with Adam, before activations.
Can I run Qwen2.5-VL-32B-Instruct on one GPU?
Yes. In 16-bit it needs about 80 GB, which fits on one RTX PRO 6000.
What does it cost to run Qwen2.5-VL-32B-Instruct?
In 16-bit it needs one RTX PRO 6000; in 4-bit, one RTX 4090 or RTX 3090. Each setup on this page shows the lowest price per GPU-hour on Arrakis right now, and you pay by the hour from a prepaid balance.
How do I deploy Qwen2.5-VL-32B-Instruct on Arrakis?
Open Models in the Arrakis console, paste Qwen/Qwen2.5-VL-32B-Instruct, choose vLLM, SGLang or llama.cpp, and rent a GPU it fits on. You get an OpenAI-compatible endpoint with its own API key, and root SSH to the machine.