Generate video with matching sound from text or a photo

Made by LTX.io

LTX-2.5 is a 22-billion-parameter open-weights model from Lightricks that generates video and sound together: describe a scene in words, or hand it a photo, and it produces a clip with a soundtrack in one pass, so picture and audio stay in sync.

New in this version are multi-shot clips — two to four connected shots in one generation — and upscalers that push the result toward 4K. It runs on your own hardware through ComfyUI or Python, but the files are tens of gigabytes and need a powerful graphics card. The weights require a free Hugging Face account and accepting Lightricks’ community license, which restricts commercial use for larger companies.

Can I use this?

Needs a powerful GPU and a ComfyUI or Python setup

You'll need A powerful GPU, ComfyUI or Python, and tens of gigabytes of disk · Setup needed

Worth knowing The weights are gated: you need a free Hugging Face account and must accept Lightricks’ community license before downloading, and that license restricts commercial use for larger companies — read it before business use. Speed and quality claims are Lightricks’ and the community’s own. BuildTube has not run or verified this model.

Why it's here

5,805 people have liked it on Hugging Face.
It was downloaded 1,588,619 times in the last 30 days.

Numbers from the snapshot taken 1 October 2026; not refreshed since.

Behind it

See the code on Hugging Face

BuildTube has not run or verified this project. Everything above is written from what the creator published.

Made this?