Nano Banana 2.1
Nano Banana 2.1 is Google's AI image generation and editing model. It keeps Nano Banana 2's speed and $0.08 price, and adds better visual design, mask-based editing and stronger subject consistency, at up to 4K.
Every image on this page was generated with Nano Banana 2.1 on fal.
What can Nano Banana 2.1 do?
The same person in every AI image
Give Nano Banana 2.1 one portrait and it puts that exact person into any scene, call after call.
- Each scene is a separate edit call with only the portrait as reference
- Face, moustache, gold hoop and linen jacket carry into every frame
- Up to 4 characters and 10 objects stay consistent, per Google
1
2
3
4
5
6
Combine up to 14 reference images in one AI image
Pass people, pets, products and props as separate references and Nano Banana 2.1 composes them into one scene.
- Six references in: a portrait, a dog, a Land Rover, a hat, a basket and a kite
- Each one keeps its look, down to the hat ribbon and the kite's stripes
- Up to 14 references per request, with 4 characters and 10 objects held consistent
Edit one region of an AI image, leave the rest
Mark the area to change and describe the change, and Nano Banana 2.1 confines the edit to it, one of Google's headline upgrades in 2.1.
- A hot-air balloon was added inside the marked box, and the box removed
- The model, the car and their reflections stay exactly where they were
- No separate mask parameter: the marked image is the input
Change the time of day in a photo
Describe the light you want and Nano Banana 2.1 relights the scene while the place stays exactly where it was.
- Midday flat light to golden hour, with the café lights switched on
- Buildings, tiles, tables and camera angle are unchanged
- Drag the slider to compare
Turn a pencil sketch into a photograph
Give it a rough drawing and a few words on materials and light, and the composition comes back as a finished render.
- The cabin, deck, stairs, jetty and rowing boat land where they were drawn
- Materials, mist and interior glow come from one sentence
- Fast concept renders for architecture, product and set design
Restyle a photo without losing the shot
Swap the medium, keep the picture: the same pose, car and field, now sculpted in clay.
- Composition, pose and palette hold through the change of medium
- Claymation, watercolour, anime or pixel art from a single prompt
- Useful for campaign variants and storyboard looks
Put a product in a new scene, label intact
Move a product shot into any setting and the packaging artwork and type stay exactly as they were.
- The tin's illustration, serif label and small print are preserved
- Lighting and shadows are rebuilt for the new scene
- One product photo becomes a season of lifestyle shots
How to use Nano Banana 2.1
Three steps from an idea to a finished image, in the playground or through the API.
What are the examples of Nano Banana 2.1?
Lush valleys, salt flats and beyond, each with the prompt that produced it.
Fashion on a mirror
"A fashion photograph on a salt flat covered by a thin layer of water that turns it into a perfect mirror: a model in a flowing butter-yellow silk gown walks toward the camera, her full reflection beneath her, the sky and soft clouds mirrored to the horizon, a vintage teal car small in the distance. Shot on 35mm film, soft high-key natural daylight from a low sun, pastel palette of teal, cream, butter yellow and rust, gentle halation and fine grain, whimsical surreal fashion-film art direction, calm and dreamlike."
Scale play
"A surreal photograph of a lush wildflower valley where a giant woman in a cream linen dress lies asleep across the rolling green hills like part of the landscape, her auburn hair spreading into the flowers, tiny hikers and a vintage sage-green Land Rover on a dirt track winding past her shoulder for scale. Shot on 35mm film, soft high-key natural daylight from a low sun, pastel palette of teal, cream, butter yellow and rust, gentle halation and fine grain, whimsical surreal fashion-film art direction, calm and dreamlike."
Surreal photography
"A tiny-planet fisheye photograph: a couple in a rust suit and a cream dress kiss on a picnic blanket on top of a small round world made of white hexagonal salt crust, a vintage teal car and a cluster of pastel balloons curving around the edge of the little planet, pale teal sky all around, the sun flaring at the top. Shot on 35mm film, soft high-key natural daylight from a low sun, pastel palette of teal, cream, butter yellow and rust, gentle halation and fine grain, whimsical surreal fashion-film art direction, calm and dreamlike."
Fashion editorial
"A fashion editorial photograph of a model in an oversized rust linen suit and a cream silk shirt, sitting on the roof of a vintage pale-blue Citroën parked in a lavender field in Provence, golden-hour light, long shadows, 35mm film grain, Vogue editorial style."
What Nano Banana 2.1 costs
Billing is per image and priced per resolution, with no minimums and no subscription. Text to image and edit cost the same, and 2.1 is priced exactly like Nano Banana 2.
| Resolution | Nano Banana 2.1 | Nano Banana 2 | Nano Banana Pro |
|---|---|---|---|
| 0.5K | Not offered | $0.06 / image | Not offered |
| 1K | $0.08 / image | $0.08 / image | $0.15 / image |
| 2K | $0.12 / image | $0.12 / image | $0.15 / image |
| 4K | $0.16 / image | $0.16 / image | $0.30 / image |
Web search adds $0.015 per request on every model, and high thinking adds $0.002 on Nano Banana 2 and 2.1. All three take up to 14 reference images. Each endpoint carries its current rate on its model page. Information updated as of October 6, 2026.
The two Nano Banana 2.1 endpoints
Text to image for new work, edit for everything that starts from an image. Same model, same parameters, same price.
Nano Banana 2.1 by Google generates images from text prompts, with output resolutions up to 4K, adjustable aspect ratios, and optional web search grounding.
Nano Banana 2.1 Edit by Google transforms reference images using text instructions, with output resolutions up to 4K and optional video or PDF context.
How to call the Nano Banana 2.1 API
The client handles the queue protocol, polls for status and returns the result when the request completes. Both endpoints take the same parameters, and edit adds image_urls.
import { fal } from "@fal-ai/client";
const result = await fal.subscribe("google/nano-banana-2.1", {
input: {
prompt:
"A fashion photo on a mirrored salt flat: a model in a flowing butter-yellow silk gown walks toward the camera, her reflection beneath her",
aspect_ratio: "3:4",
resolution: "2K",
},
logs: true,
onQueueUpdate: (update) => {
if (update.status === "IN_PROGRESS") {
update.logs.map((log) => log.message).forEach(console.log);
}
},
});
console.log(result.data);
console.log(result.requestId);
Nano Banana 2.1 FAQ
What is Nano Banana 2.1?
Nano Banana 2.1 is Google's AI image generation and editing model, released on October 6, 2026. It keeps the Flash-level speed and cost of Nano Banana 2 while improving visual design, mask-based editing, subject consistency, text rendering and very wide or tall aspect ratios. On fal it is available as two endpoints: text to image and edit.
What is new in Nano Banana 2.1 compared to Nano Banana 2?
Google highlights three big gains: more polished visual design, mask-based editing that changes only a selected region, and stronger subject consistency across edits and related images. It also fixes tiling artifacts at 4:1, 8:1, 1:4 and 1:8 at 2K and 4K, and improves text and infographic accuracy. The price is unchanged. The one thing that goes is the 0.5K draft resolution, which stays available on Nano Banana 2.
How much does Nano Banana 2.1 cost?
$0.08 per image at 1K, the default. 2K costs 1.5x ($0.12) and 4K costs 2x ($0.16). Turning on web search adds $0.015 per request, and high thinking adds $0.002. Text to image and edit are priced the same, and there is no subscription.
How does mask-based editing work in Nano Banana 2.1?
Send the image to the edit endpoint with the region you want changed clearly marked, for example a coloured box drawn over it, and say in the prompt to edit only inside the mark and remove the mark. The example on this page added a hot-air balloon inside a magenta box and left the model, the car and their reflections in place. The rest of the frame is regenerated rather than copied bit for bit, so expect it to match closely rather than pixel for pixel.
How many reference images can I use?
Up to 14 per request, according to Google, with consistency for up to 4 characters and 10 objects. Pass them in image_urls and refer to them by order in the prompt ("the woman from the first image", "the cap from the second image"). The edit endpoint also accepts an optional video, audio file or PDF, up to 15 MB each, as extra context.
What resolutions and aspect ratios are supported?
Resolution is 1K, 2K or 4K. Aspect ratio can be auto, 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3 or 9:16, plus the extreme ratios 4:1, 1:4, 8:1 and 1:8. Up to 4 images per request, in PNG, JPEG or WebP.
What do web search and thinking do?
enable_web_search lets the model ground the image in current information from the web, which helps with real places, products and events. thinking_level set to minimal or high lets the model reason about the request before it renders; high is the setting to try for dense layouts and multi-step instructions.
Nano Banana 2.1 vs Nano Banana Pro: which should I use?
Start with Nano Banana 2.1: it is the newest model in the family, costs about half as much at every resolution, and carries the latest gains in design, editing and consistency. Nano Banana Pro remains available for workloads already tuned around it.
Can I use Nano Banana 2.1 for commercial projects?
Yes. Images generated through the fal API can be used in commercial projects. Check fal's terms of service for full details on usage rights and licensing.
How do I get started with the API?
Install the fal SDK for Python or JavaScript, grab an API key from your dashboard, and make your first request in a few lines of code. Serverless, with no GPUs to manage. The API documentation lists every parameter.
Get in touch about Nano Banana 2.1
Generating images at volume? Leave your details and our team will get back to you about pricing and enterprise terms.