Content Studio: Turn any topic into a faceless video.

Content Studio is DreamBench’s AI faceless video generator. Type a topic or a question and Claude writes the script scene by scene and draws and animates every scene in one look, a narrator reads it with burned-in captions, and you get a vertical short for YouTube Shorts, TikTok and Instagram Reels.

Type your topic. Leave the kind (What if, Explainer, Short story or Facts) and the look on Auto for Claude to pick, or pick them yourself; choose a narrator, a language and a length, check the price on the button and start. Claude writes the script, draws and animates each scene, a narrator reads it, and you download the finished MP4.

Yes. Every scene is animated by Claude in code: Claude writes a small program for each scene against the look’s tested kit — its characters, places, labels and captions — and DreamBench renders it frame by frame. Claude checks a few frames of each scene and fixes what is wrong before the episode is made, and no picture or video model is used. There are 12 looks: Toy Town, Voxel World, Data HUD, Botanical Plate, Swiss Grid, Pixel Quest, Paper Cut-out, Watercolor, Anime, Noir Comic, Vintage Poster and Clay.

Content Studio is priced by the episode, in the credits of your DreamBench plan. An episode of 30 seconds costs 838 credits and an episode of 1 minute costs 1,466 credits, whatever the look. The price is on the button before anything is made, each episode is its own charge, and an episode that fails gives its credits back.

Yes, for making the videos. One run makes up to 4 episodes written as a series, each one standing alone, and every episode comes with a title, a post caption and hashtags. You download the MP4s and post them yourself: Content Studio doesn’t post or schedule to YouTube, TikTok or Instagram.

Yes. Open a finished episode and press Edit: change any line, what a scene shows or how it moves, the narrator, the look or its theme. Only what you change is made again, at the price on the button — reading one line again costs 19 credits and drawing one scene again 139 credits — and your video stays as it is until the new one is ready.

Every video is a vertical 9:16 short at 1080 × 1920 and 30 frames a second, as an MP4. An episode runs 30 seconds, 1 minute, 2 minutes, 3 minutes or 5 minutes.

There are 8 narrators, all ElevenLabs voices: George, Brian, Matilda, Lily, Daniel, Jessica, Will and Bill. Each one narrates in 8 languages: English, Brazilian Portuguese, German, Japanese, Simplified Chinese, Indonesian, Spanish and French. The captions are in the narration’s language, burned into the video a few words at a time and timed to the voice.

No. Nobody presents a Content Studio video: the scenes are drawn and animated by Claude, and the script is read by an AI narrator. That also means you can’t upload your own voice.

No logo is laid over the video itself. With Starter, each episode ends with a short DreamBench.ai card after its last scene; with Plus, Ultra and Studio, it ends on the last scene.

No. The script is Claude’s best knowledge of your topic, and it can be wrong or out of date. Read the narration and check its claims before you post, as you would any draft.

It adds no background music, doesn’t post or schedule to TikTok, YouTube or Instagram, and doesn’t start from a link to someone else’s video. It is made inside DreamBench: there is no connector for Claude or other assistants.

Claude writes the script and draws and animates every scene in code, with no picture or video model. An ElevenLabs voice narrates every episode.

Yes. Every file Content Studio makes is marked as AI-made, as all DreamBench output is. When you post, follow the platform’s own rules for labeling AI-made videos.