For Shorts creators
Generate a vertical Short from one idea: AI visuals, voiceover, and word-by-word captions in 9:16. The draft lands on a real timeline, so you can re-cut the first two seconds until the hook actually hooks.
Vertical 9:16 output sized for Shorts, TikTok, and Reels.
AI visuals, voiceover, and synced captions from one prompt.
Re-cut the hook without regenerating the rest of the video.
GPU rendering in the browser, no upload queue.
Anyone can generate a Short. The channels that grow are the ones that iterate: testing hooks, tightening pacing, and adjusting captions based on what retains. A locked AI export cannot iterate. FrameCompose gives you the generated draft and the editor in the same place.
When a Short underperforms, open the project, change the first two seconds, and export again. No credits, no full regeneration.
Scenes generate in 9:16 with captions positioned for phone screens and safe areas respected. You are not cropping a landscape video after the fact; the Short is vertical from the first frame.
Most Shorts are watched with sound off at least at first. Word-by-word captions generate synced to the narration, and every word stays editable. Pick fonts and styles that match the channel, and reuse them across videos.
Everything runs in the browser: generation, editing, and hardware-accelerated rendering with WebGPU and WebCodecs. Scenes cost about 3 cents to generate and edits are free, so a daily schedule stays affordable.
A few quick answers before you jump in.
Type an idea or script into Faceless Studio. It generates the scenes, voiceover, and captions in vertical 9:16, then drops everything on an editable timeline. Adjust the hook, pacing, and captions, then export directly for Shorts.
The first two seconds. Shorts are swiped away instantly, so the hook has to land immediately. Because FrameCompose output stays editable, you can re-cut the opening or reword the first caption without regenerating the video.
Yes. The 9:16 export works across YouTube Shorts, TikTok, and Instagram Reels. One project can be exported multiple times as you iterate.
Yes. Word-by-word captions generate automatically, synced to the voiceover, and stay editable down to the word with fonts and styles that read well on a phone.
Rendering runs in your browser on the GPU using WebGPU and WebCodecs, at native speed. There is no upload queue, so iterating on a Short takes seconds, not minutes.
Explore the rest of the workflow, from first idea to final export.
The full Faceless Studio workflow: idea to finished video without a camera.
Read pagePaste a Reddit thread and turn it into a narrated, captioned story video.
Read page100+ free transitions and effects to keep Shorts moving between scenes.
Read page