3D animatic for AI video generation
Show AI how your shot should move. Build a simple 3D scene, animate the camera and objects, then use the animatic as a video reference in Seedance and other models that accept video.
Sign in to Neylora to get started. The AI agent uses credits.
What is an animatic, and why use one for AI video?
An animatic is a draft video that already shows composition, movement, and timing. It helps you test an idea in motion: how the camera approaches the subject, when a prop appears, and how long each shot lasts.
“The camera orbits the car” leaves the direction, height, and speed open. In a 3D animatic, you choose them yourself. It gives generation a clear brief: staging and movement from the video, appearance from the prompt and images.
What you can define before running AI
Show the camera movement you want
Push in on a product, orbit a car, or fly through a room — set a path in 3D and view it through the camera. The video reference shows the model the direction and speed of movement.
Compose your shot
Arrange objects and choose framing and angle. Check that the subject is not hidden by the set and the product stays in frame throughout the movement.
Set the order and timing of actions
Place animation keys on the timeline: when an object starts moving, where it stops, and when the shot changes. Preview the scene's rhythm before generating.
Start with a description or image
Ask the AI assistant to build a simple scene from your idea. Then refine object and camera positions manually — clear shapes and movement are enough for a draft.
Refine the staging before generation
Compare scene variations in one project. Spot an awkward angle or a camera move that's too fast in the draft, before paying to run a video model.
Use the animatic as a video reference
Export the camera view as an MP4 directly to your Neylora reference library. Add the video to a supported model and describe the final shot you want.
How to create an animatic and use it in Seedance
Build a scene
Block out your subject, props, and environment using simple 3D shapes. Their size and position matter most at this stage; appearance details can be added during generation.
Animate the scene
Choose an angle, add camera and object keyframes, and set the duration. Preview it and check that viewers have time to see what matters.
Export a video reference
Click Export to save the animatic as an MP4 in your references. Match the duration and aspect ratio to your chosen video model's requirements.
Add style and generate
Attach the video in the generator. Specify which movement to preserve, then describe appearance, materials, lighting, and atmosphere in the prompt or provide image references.
Example: a car commercial
You need an eight-second shot: the camera starts by the front wheel, pulls back, and reveals the car from the side. Block out the car with a simple shape, set the camera's start and end positions, and preview the entire move.
Use this video as the motion reference and a photo of the car as the appearance reference. Specify body color, environment, and lighting in the prompt. Agree on the staging first, then use generation attempts to refine image quality and detail.
The same approach works for product orbits, interior walkthroughs, a subject entering the frame, or a sequence of shots for a music video.
Animatic FAQ
How is an animatic different from a storyboard?
A storyboard shows individual shots; an animatic adds timing: how long shots last, how the camera and objects move, and when cuts happen. A 3D animatic lets you adjust staging and immediately preview the motion.
How do I use an animatic to generate video in Seedance?
Build a scene in Neylora, animate it, and click Export to add an MP4 to your video references. Choose a Seedance mode supporting video references and attach it. Specify what to take from the animatic: camera movement, object placement, or action order. Define appearance and style separately with text and images.
Can I use an animatic with other AI models?
Yes, if the selected mode accepts video as a motion reference or an editing source. For example, Runway Aleph works with source video. For text- or image-only modes, use the animatic as a plan for starting frames and scene descriptions. Video support and limits depend on the model and mode.
Will AI reproduce the animatic's movement exactly?
A video reference provides guidance, not a frame-by-frame guarantee. The model may change the path, speed, or scene details. Start with a short, clear movement, specify what to preserve, and compare the result with the animatic.
Do I need to prepare 3D models beforehand?
No. Built-in primitives are enough for an initial test. For a specific silhouette, adjust the shape using geometry tools or ask the AI assistant to build a starting point.
Can the AI assistant build a scene from a photo?
You can use a photo as a guide for a simplified 3D scene. The assistant creates a starting point you can refine manually. This is not precise 3D reconstruction: check hidden object parts, dimensions, and distances.
Do I need credits to work on an animatic?
Placing objects, adjusting the camera, and animating manually do not require the AI assistant. The assistant uses credits: you need at least 5 available credits to start or continue an AI task, and actual usage is shown in chat. Final AI video generation is charged separately according to the model and settings.
Start with one movement
Place an object, choose an angle, and animate the camera. Preview the scene, then send your draft to references for your AI video.