README.md

metadata

title: LTX Video Generator
emoji: 🎬
colorFrom: purple
colorTo: pink
sdk: gradio
sdk_version: 6.20.0
app_file: app.py
pinned: false

LTX Video Generator (Personal Use)

A free text-to-video generator built on LTX-Video, running as a Hugging Face Space with a Gradio UI.

Deploy steps

1. Go to https://huggingface.co/new-space
2. Fill in:
   • Space name: e.g. my-ltx-video-generator
   • SDK: Gradio
   • Hardware:
     • Free CPU works but will be very slow (many minutes per clip).
     • ZeroGPU is free if your account is eligible — best option for personal use.
     • Or pick a paid GPU tier (e.g. T4/A10G) billed by the minute if you want faster, always-on generation.
   • Visibility: Private (recommended, since this is personal use) or Public.
3. Once the Space is created, upload these three files (or push via git — see below):
   • app.py
   • requirements.txt
   • README.md
4. The Space will build automatically. First build takes a while — it has to download the LTX-Video model weights (several GB).
5. Once it says "Running", open the app and generate your first clip.

Uploading via git (alternative to the web UI)

git clone https://huggingface.co/spaces/YOUR_USERNAME/my-ltx-video-generator
cd my-ltx-video-generator
# copy in app.py, requirements.txt, README.md
git add .
git commit -m "Initial LTX video generator"
git push

You'll need a Hugging Face access token (with "write" scope) to push: Settings → Access Tokens → New token, then use it as your password when git prompts for credentials (username = your HF username).

Notes

• LTX-Video is gated/openly licensed under its own model license — the first time the Space downloads it, you may need to have accepted the license on the model page with the same account that owns the Space.
• Generation settings (resolution, frame count, steps) are exposed in the UI so you can trade off speed vs. quality.
• This is set up for personal/private use — keep the Space private if you don't want to share generations publicly.