Wan2GP

Super Optimized Gradio UI for Wan2.1 video generation, designed for GPU-poor machines (5GB+ VRAM). Generate up to 12 second AI videos with text-to-video and image-to-video modes.

Based on Wan2GP by deepbeepmeep.

Requirements

• NVIDIA GPU with 5GB+ VRAM
• Windows or Linux

How to Use

1. Click Install to set up the environment (clones the repository, installs dependencies, PyTorch, triton, and sageattention).
2. Click Start to launch the Gradio web UI.
3. The web UI will open automatically once the server is ready.
4. Use Advanced > Compiled mode for faster inference (may not work on all systems).
5. Place LoRA files in the T2V Loras or I2V Loras folders accessible from the sidebar.

Features

• Text-to-Video generation (up to 12 seconds)
• Image-to-Video generation
• Multiple image input support
• LoRA support (T2V and I2V)
• Optional compiled mode for faster inference
• Optimized for low VRAM GPUs (5GB+)

API

Wan2GP exposes a Gradio web interface. You can interact with it programmatically:

Python

from gradio_client import Client

client = Client("http://127.0.0.1:7860")
result = client.predict(
    prompt="A cat walking on the beach at sunset",
    api_name="/generate"
)
print(result)

JavaScript

import { Client } from "@gradio/client";

const client = await Client.connect("http://127.0.0.1:7860");
const result = await client.predict("/generate", {
  prompt: "A cat walking on the beach at sunset",
});
console.log(result.data);

cURL

curl -X POST http://127.0.0.1:7860/api/generate \
  -H "Content-Type: application/json" \
  -d '{"data": ["A cat walking on the beach at sunset"]}'

Note: The port may vary. Check the Pinokio terminal output for the actual URL. The API endpoints depend on the Gradio interface exposed by the app.