A demo for SulphurAI/Sulphur-2-base, a video model fine-tuned from LTX-2.3 (22B) that generates short clips with synchronised native audio. Describe a shot for text-to-video, or upload a still to animate it.
Give a descriptive prompt and generate a short clip with matching audio.
Upload a starting frame and let the model bring it to life.
Optionally expand a short prompt with the built-in Gemma-3-12B encoder.
app.py) and is written for
Hugging Face ZeroGPU. It needs GPU hardware to run, which this account can't host
yet — a community GPU grant has been requested. Once approved, this page switches to the
live demo.