AI studio
AI Text-to-Video Generator
AI Studio
Turn a text prompt into a short cinematic video clip, powered by state-of-the-art generative AI.
Usage & publishing rights — please read
- The video you generate is for personal use only by default.
- You may not upload or monetize it on YouTube or any other platform unless you purchase the required YouTube Upload pass (or an eligible plan).
- After purchasing, you get the rights to publish to your own channel and use the video commercially, according to your selected plan.
Advanced
Describe what you want to generate to continue.
Your generated video will appear here.
Describe what you want, hit generate, and the render runs on our servers — you can close the tab.
Questions answered
Frequently asked questions
Yes — the clip is rendered on our servers, so you have to be signed in, and each render spends credits (8 at the current price, shown live on the badge at the top of the page). If the render fails, the credits are refunded automatically.
Pick a length from the Duration dropdown — currently up to 10 seconds. That list is built from what the video engine on the server can actually render, so it only ever offers durations it can serve rather than promising a length it would have to fake.
A prompt of up to 2,000 characters, a resolution from a fixed list (1280×720, 720×1280, 1024×1024, 832×480 and 480×832), the duration, and — under Advanced — a negative prompt and a seed, where 0 means random and any other number makes the result reproducible. There are no model, step-count or guidance settings.
The clip comes back as a silent MP4 — text-to-video generates pictures, not audio. LrcSong runs the model on its own hardware rather than a third-party API, and when the built-in engine is serving this tool the page shows a heads-up saying so: clips are short and rendered small, then frame-interpolated and upscaled before download. Asking for a longer duration gives you the full length with slower motion rather than more distinct action.