A red fox trots through a snowy pine forest at golden hour, camera tracking alongside, paws crunching in the snow.
Official demo · live on ZeroGPU
Direct the whole scene.
Turn text or a starting image into cinematic video with synchronized sound. LTX-2.5 keeps motion, voice, ambience, camera work, and visual style in one generation.
LTX / WORLD MODEL / 02.5
Start with direction.
Strong prompts specify subject, motion, setting, camera, light, and sound. Copy one, then shape it into your own shot.
A clockwork hummingbird hovers beside a blue orchid, tiny gears whirring, wings beating rapidly, soft workshop ambience.
A lone astronaut watches dawn spread across Mars. Slow push-in. She whispers, “Tell Earth we made it.” Wind brushes the microphone.
Generate a shot.
The embedded app is Lightricks' official LTX-2.5 Space. ZeroGPU availability and daily usage quotas are managed by Hugging Face.
Data note: prompts and optional images entered inside the embedded demo are sent directly to Lightricks' official Hugging Face Space for generation. This presentation Space does not receive or store them.
22Bdistilled video transformer
8 + 3two-stage inference steps
24 fpsgenerated video cadence
A/Vsynchronized picture and sound