AI movie maker: from a logline to a finished cut

Scenaloom's video projects take a film from one sentence to a finished cut in the browser. You say what the film is about, Scenaloom plans it as short reels with their own shot lists, you draw a frame and shoot a take for each reel with an AI video model, and the takes come together on a timeline you can play, share and export. The price of every frame and take is on its button before you press it.

How the AI movie maker works

1. Start a video. Pick the frame shape (16:9 for widescreen, 9:16 for vertical), the type (drama or commercial), the camera (steady or handheld) and one of twelve looks, from 1940s noir to cel animation. Add a logline: one sentence on what the film is about.
2. Plan the reels. Scenaloom splits the film into reels of 5 to 15 seconds, each with one to four timed shots that say what the camera sees and hears. Keep, edit or rewrite them before anything is generated.
3. Draw a frame. A storyboard frame for a reel shows what the take will look like, so you can fix framing or casting while it costs a frame rather than a video.
4. Shoot the take. A video model renders the reel from its shot list, look and direction. A reel can continue from the one before it, so the action carries across the cut.
5. Cut the film. The timeline puts the takes in order with audio tracks and subtitles. Play it through, share a link, and export the film as a video file.

Keeping characters and places the same

The hard part of an AI film is not one good clip; it is the tenth clip still looking like the first. In Scenaloom the cast and locations live in the project's story bible, with their pictures. Each reel names who is in it and where it happens, and models that take reference images receive those pictures with every take, along with the film's look. Models that work from text only get the descriptions, and the reel says so before you shoot.

What it costs

Frames, takes and narration use credits, and the price of each run is shown on its button before you start. It depends on the model and the length of the clip. If a model fails to finish a take, the credits go back to your balance. Free accounts get daily refresh credits and a daily spending limit; see plans and pricing.

Vertical films and short series

Choose 9:16 when you create the video and every frame and take is drawn vertical, for phone screens and short-form platforms. For a series, write the episodes as a script project first: the same cast and places then carry from episode to episode.

Script to video AI: how a script becomes shots, frames and video

If you already have a screenplay, Scenaloom turns it into video scene by scene, and you approve each step before anything is generated:

1. Write or import the script. Scene headings become scenes, and character cues and scene headings become the cast and locations of the project's story bible. Give the main characters and places a picture there.
2. Fill reels from the script. In Production, each scene is drafted into reels of 5 to 15 seconds, and Suggest shots proposes the timed shots for a reel. Each reel names its cast, its location and its direction for mood, lens and light; edit any of it first.
3. Draw the frames. A storyboard frame shows what each reel's take will look like, drawn with the cast and location pictures, so you fix framing while it costs a frame rather than a video.
4. Shoot the takes. A video model renders each reel from its frame, shots and direction. Shoot again until you have the take you want, and pick it as the final take.
5. Cut the film. The chosen takes play in order on the timeline with audio and subtitle tracks, and the film exports as a video file from your browser.

If you are starting from an idea instead, the logline is enough to plan the first reels. Either way the story stays in one project, so the next take starts from the decisions the film has already made.

Start a video or read how storyboards and production work.