Storyboard Generator
Script to visual storyboard
On it. Running Storyboard Generator now and I'll bring the results straight back here.
Try asking
About
Storyboard Generator turns an enhanced script into a visual storyboard — one AI-generated frame per shot, composited into a labelled grid image and saved to your library, ready to hand straight to video generation.
It's the bridge between scripting and production. Once you've enhanced a script, this tool generates a realistic first-frame image for every shot in that script, applies your style reference and persona for visual consistency, then stitches everything into a grid so you can review the whole video at a glance before spending a penny on video generation. Each frame is also saved individually with its video generation prompt attached.
What you can do
- generate_storyboard — generate one AI frame per shot from an enhanced script, composited into a storyboard grid image
Who it's for
Marketers, DTC brands, and content agencies who want to see a video before they commit to generating it. Works best as part of the full UGC or video production pipeline.
How to use it
- Run the script-enhancer tool first to produce an enhanced script file
- Call generate_storyboard with the script file ID and a style reference — either a saved style reference ID or an inline description like "super 8, grainy, domestic realism"
- Optionally pass persona, scene, product, or outfit file IDs for visual consistency across shots
- Review the returned storyboard grid — show the URL to the user
- Use file_read with the returned file ID to get per-shot frame URLs and video generation prompts, then pass those to your video generation tool
Getting started
You need an enhanced script file from the script-enhancer tool and at least one style reference. Upload persona and product references at your file library to get the best results.
Generates one AI frame per shot from an enhanced script file and composites them into a storyboard grid image saved to your library. Style reference required. Accepts personas, scenes, products, and outfits. ⏱ Takes 1–3 minutes.
- Saved personas and scenes are now labelled for the model automatically, so frames hold identity and location without extra prompt wording.
- Shot video now defaults to Seedance 2.5.
- Initial release