corehu
Rent a GPU

USE CASE

Generate AI videos on a cloud GPU

Make text-to-video and image-to-video clips with open models like Wan 2.2 in ComfyUI on an RTX 5090, A100 or H100. Billed per second, no subscription, no render queue and no watermark.

See RTX 5090 machines Checking live prices…

Billed per second from your USD balance. Stop or terminate the machine when you are done.

Recommended GPU

RTX 5090 · A100 SXM4 · H100 SXM

Video models are the most VRAM-hungry AI jobs. Wan 2.2 5B runs comfortably on a 32 GB RTX 5090; the 14B models, longer clips and higher resolutions are faster on an 80 GB A100 or H100.

Template to pick

PyTorch

NVIDIA CUDA + PyTorch development environment, ready with SSH and Jupyter.

Up and running in four steps

  1. 1

    Rent the machine

    Open the catalog with RTX 5090 and the PyTorch template already selected and press Rent.

    See RTX 5090 machines
  2. 2

    Connect with SSH

    Copy the SSH command from your machine's page and add the port forward for ComfyUI.

    ssh -p <PORT> root@<HOST> -L 8188:localhost:8188

    The SSH command is on your machine's page. The -L part forwards the port so the app opens in your own browser.

  3. 3

    Install ComfyUI and the Wan 2.2 model

    Run these commands once. They install ComfyUI, download the Wan 2.2 5B video model, its VAE and text encoder into the right folders, and start ComfyUI.

    apt-get update && apt-get install -y wget
git clone https://github.com/comfyanonymous/ComfyUI
cd ComfyUI && pip install -r requirements.txt
wget -P models/diffusion_models https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/resolve/main/split_files/diffusion_models/wan2.2_ti2v_5B_fp16.safetensors
wget -P models/vae https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged/resolve/main/split_files/vae/wan2.2_vae.safetensors
wget -P models/text_encoders https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/resolve/main/split_files/text_encoders/umt5_xxl_fp8_e4m3fn_scaled.safetensors
python main.py --listen 0.0.0.0 --port 8188
  4. 4

    Write a prompt and render

    Open http://localhost:8188, go to Workflow → Browse Templates → Video and pick Wan 2.2 5B. Type your prompt (or add a start image) and press Run. Copy finished videos to your computer with this command.

    scp -P <PORT> -r root@<HOST>:/root/ComfyUI/output ./videos

Questions

How can I create a video with AI?

Rent a GPU, install ComfyUI, download an open video model such as Wan 2.2 and describe the scene in a prompt, or start from an image. The steps above take about 15 minutes the first time; each clip then renders in a few minutes.

Which GPU do I need for AI video?

An RTX 5090 (32 GB) for Wan 2.2 5B and short 720p clips. For the 14B models, longer clips or batches pick an A100 80 GB or H100.

Can I use the videos commercially?

That depends on the model's license, not on CoreHu. Wan 2.2 is released under Apache 2.0; check the license of every model and LoRA you download.

How much does it cost?

You pay per second at the hourly price shown above, only while the machine runs. There are no per-video credits. Stop or terminate the machine when you are done; a stopped machine still bills for its disk.