Can you make an animation if you can't draw?
Yes — because in a text-to-animation workflow, nothing depends on your hands. The character already exists as an animated stickman rig; the poses, in-between frames, timing, and camera moves are generated from your written description. Your job shifts from drawing to directing: deciding what happens, in what order, and how it should feel. That's a writing skill, and everyone who can describe a scene to a friend already has it.
The other half of the answer is just as important: what you get back is not a locked video file. The generation lands as editable clips on a timeline, so when something looks off you fix that one clip instead of starting over. Fixing is also drawing-free — you drag clip edges, swap expressions, or nudge a pose with your mouse.
Step 1 — Describe the scene like you'd tell it to a friend
Open the studio and you're in the Scene panel with a prompt box. Write the action as beats, not keywords:
Every beat in that sentence — the walk, the watch checks, the jump, the run — becomes its own motion clip on the timeline. Direction words ("from the left"), pacing words ("slowly", "suddenly"), and camera wishes ("zoom in on his face") all translate into real structure. Vague one-word prompts like "funny" give the generator nothing to build beats from; a short story does.
Step 2 — Generate and look at the timeline, not just the canvas
Press Generate and the scene plays. The habit that separates people who get good results from people who re-roll forever: look below the canvas at the timeline editor. Each phase of your prompt is a clip you can trim, stretch, reorder, or delete. If the jump reads too slow, you shorten that clip — the scene rebuilds with smooth transitions automatically. You never regenerate the whole thing because one second was wrong.
Step 3 — Dress the scene without drawing anything
Everything visual comes from pickers, not pencils. The sidebar's Character category has outfits, headwear, and handheld props; Backgrounds has full scenes with a search box; Particles adds rain, sparks, or impact effects. Faces are handled by an expression system — pick a look and it rides the motion on its own track. And if you'd rather let the AI decorate, switch the AI detail setting from Lite to Assisted (adds a background) or Full (adds wearables and props too), then swap out anything you disagree with.
The one place drawing can enter — entirely optionally — is custom art: you can upload an image or doodle and attach it to a body part, and it moves with every animation you generate.
Step 4 — Add a voice by typing it
If the character should speak, you don't record anything. Type the line, and an AI voice speaks it synced to the motion, with subtitles generated automatically. Performance tags like [pause] and <excited>…</excited> shape the delivery. The voice-over guide covers this end to end, and Dialogue mode handles two characters talking.
What does it cost to start?
New accounts start with a free token balance, and generating costs tokens based on the characters (letters) you type — spaces are free. A typical scene prompt is a sentence or two, so the starting balance covers several complete animations with room to experiment. Editing on the timeline costs nothing; you only spend tokens when you generate.
The whole flow in five steps
- Open the studio and write the scene as two or three beats in the prompt box.
- Pick an AI detail level — Lite if you'll decorate yourself, Full for a one-prompt scene.
- Generate, then trim or reorder clips on the timeline until the timing reads right.
- Swap outfits, background, and expressions from the sidebar pickers.
- Add a typed voice line if the scene needs one, then export.
If you want to understand the text-to-animation engine itself — what kinds of prompts work, what the AI detail levels change, and what its limits are — read how to turn text into animation with AI next.