Edit images by asking
Describe the change in plain words and Nano Banana rewrites only that, leaving the rest alone.
Start creating
What makes Nano Banana different
The everyday tier: it edits as well as it generates, and it is quick enough to keep trying.
Edits you can describe
Ask for the change in plain words. Everything else stays put.
Subjects that survive
Faces, logos, and packaging hold their shape edit after edit.
Quick enough to explore
Renders land in seconds, so trying ten directions is normal.
Stills that become shots
Hypernatural hands any frame to a video model as the shot's opening frame.
Change one thing, not the whole image
Swap the background, recolor the jacket, lose the extra hand. Nano Banana touches what you asked about and hands back the rest of the frame unchanged, so an edit never costs you the version you already liked.
Product shots that keep the label
Upload the real packaging once and your product comes back as itself — the wordmark, the colors, the shape — in every scene you build around it. Add your products and characters to Hypernatural's library and @mention them shot to shot.
A whole board in an afternoon
Fast renders mean you can lay out every shot in a sequence, look at it as a whole, and re-roll only the frames that are not working — instead of defending the first image because it took so long.
FAQs
Everyday image work where you expect to revise: product shots, social frames, and scene stills you will tweak a few times before they are right. It is the tier to reach for when the job is a change to an existing image rather than a brand new one.
Nano Banana is the fast everyday tier; Pro is the top-fidelity one, built for legible text and crowded compositions. In Hypernatural you can switch models and re-render one shot at a time, so you can rough a sequence out here and re-render only the frames that ship on Pro.
Yes. Upload a photo, a product shot, or a previous render as a reference and describe what should be different. Nano Banana works from your image rather than starting a new one from scratch.
Yes. Create the character once in Hypernatural's library, then write its @handle in any shot — the @ is what attaches the portrait, so a bare name will not do it. They come back as themselves, and the portrait carries over if you later switch models and re-render the shot. What they wear follows the shot.
Yes. Hypernatural hands any still to the video model you pick, then captions, narrates, and cuts it together with the rest of your shots — all in the same editor, no exporting in between.
No. Describe what you want in ordinary language and Hypernatural writes the image prompt for you. You can open it and edit the wording yourself whenever you want more control.
Hypernatural Desktop
Make videos with Nano Banana from Claude Code or Codex
Hypernatural Desktop has Nano Banana built in. Ask your agent to use it in any workflow, like these.
- hn/workflow-talking-head-extra-edit
Talking head extra edit
Turn a raw webcam or talking-head recording full of retakes into a tight rough cut of the best takes, then make it EXTRA in the client's own brand: kinetic titles, karaoke captions, synthesized sound effects and motion graphics that react to the speaker's face and hands, with type behind them that's never more than a little hidden. Use for founder or creator explainers, launch and feature videos, and any 'cut my takes and go wild with graphics' request.
WorkflowOfficial - hn/workflow-script-to-video
Script to video
Make a finished narrative video from a written script: adapt it into script.json, cast and perform the voices, design the sets, stage the cast in stills, compose every shot with its dialogue spoken in the cast voices, composite titles over the footage, then make the cuts and deliver. For scripted shorts, skits, training dramas, quiz shows and multi-character scenes made from nothing but text.
WorkflowOfficial - hn/workflow-product-explainer-video
Product explainer video
Make a ~35 s faceless marketing explainer for an app or product from its own code repo or website: a voiceover over kinetic type in the brand's colours and the product's real screenshots in a phone frame, with a music bed, built as one hn4a-compositor spec cued from the voice take's own word times. Problem, brand, three how-it-works steps, one feature, social proof, end card. Covers mining the repo for copy, facts and images, a script and aspect gate, the opening scene in a few looks on estimated word times shown live minutes after the script, beside the voice and music auditions, then the whole layout in the picked look cued to the picked voice, shown live before the render. Costs well under $1. For an ad with a presenter avatar and rebuilt UI, use workflow-product-ui-video.
WorkflowOfficial
Community
© 2026 Hypernatural Systems, Inc.