How it works
From idea to finished video.
Eight steps. You approve the price at every one that costs money.
Pitch the idea
Type it the way you'd say it: “30-second ad for a Denver roofer, upbeat, ends on the phone number.” The AI Producer drafts a shot list with a budget. Nothing is spent yet.
Lock your cast
Add each character. Approve one still from the cast sheet and it's locked for every shot.
Shape each shot
Edit the shot or leave a director's note. You'll see a clean v2 prompt and a diff, and you approve it.
Preview first
A 4-second, low-res take for a few credits shows whether the direction is right before you pay for the full render.
Quote, confirm, render
The quote shows seconds × model × takes. Confirm it and the full shots render, inside your project budget.
Voice it
Choose a stock voice, upload one, or record yourself in the browser. Your script becomes the voiceover.
Export and post
Shots, voice, music, captions and watermark are stitched into one 9:16 or 16:9 file, with AI-content labels in the file data.
Keep the receipts
Download the rights pack: prompts, versions, model IDs, dates, consents and charges for the whole project.
Formats and specs
| Item | Detail |
|---|---|
| Shot length | 4, 6 or 8 seconds per shot; longer videos are stitched from shots |
| Aspect ratios | 9:16 (Reels, Shorts, TikTok) and 16:9 (YouTube, web, TV) |
| Resolution | 720p previews, up to 1080p finals |
| Sound | Native sound from the video model, plus your voiceover and music bed |
| Video model | Google Veo 3.1 on Vertex AI (generally available versions) |
| Labels | Invisible SynthID watermark on every clip; AI-content metadata on exports |
Make your first video with us.
Early-access spots open in small groups. Join the list or call 1-800-481-8638.