Edit the script. The video follows.

Thothium is a faceless-video studio for Windows that turns a script into a finished video: scenes, narration, captions, timeline editing and GPU export in one app, with YouTube publishing built in. It runs as automated or as hands-on as you like, from one-click generation to scene-by-scene fine-tuning.

Yes, completely free: no card and no usage credits. Founding editors lock alpha pricing forever when we launch paid tiers.

Windows, and that is the only hard requirement. An NVIDIA card with 8 GB or more gets the fastest path: local picture painting, the richer voices, voice cloning and NVENC export. Without one, everything still works on the processor, at roughly two minutes of work per minute of finished video, with pictures coming from the cloud instead. macOS and Linux are on the roadmap. Your specs just help us prioritize who gets keys first; nothing is excluded.

No. If you can edit a doc, you can edit a video, because the script is the edit. The full timeline is there when you want it and ignorable when you don’t.

Narration, captions, playback, editing, rendering and image/video generation all run on your GPU. The internet only comes in when you want it: script research, fetching web images, or optional cloud generation. Results are cached, so the same shot is never generated twice.

Yes. Connect a channel and export straight into a publish queue: titles and descriptions are drafted for you, and uploads are spaced into daily publish slots instead of landing all at once. You choose what gets queued, and nothing ships without you.

Fully. Projects live on your disk as ordinary files, exports are yours, and nothing uploads anywhere unless you queue it for publishing or pick cloud generation for a shot.

Alpha testing has started, and keys go out in small batches so we can answer everyone who hits a problem. When yours is ready you’ll get a Reddit DM, and possibly an email. Nothing to do until then, though the Discord is open if you want to follow along.