Google’s next-generation Flash image model renders posters and signs with legible typography, keeps people consistent across shots, and edits photos from plain-language instructions — at up to 4K.
30 free credits on sign-up · Failed jobs refunded
What you can set for each generation on SeeBloom.
Where the model stands out compared with other image generators.
Movie posters, shop signs, packaging and infographics come out with clean, correctly spelled Chinese and English text — far fewer garbled letters.
Combine up to 14 reference images to keep faces, outfits and products consistent across a series of images.
Change the background, swap clothing, relight a scene or remove objects by describing the edit — no masks or layers needed.
Generate at 1K for quick ideas and 4K for print, banners and high-resolution product shots.
Copy a prompt, open the studio and paste it in — then adjust the subject and details to make it yours.
Widescreen movie poster: rainy neon old street, wet flagstones reflecting red and teal neon; an East Asian swordswoman in a long black coat stands center with her back to us, blade in hand, layered signs and mist beyond. Bold Chinese calligraphy title "夜行者" top center with a small English line "THE NIGHT WALKER" beneath. Crisp, centered, symmetrical typography, cinematic realism, crimson vs. cold teal, no watermark.
Widescreen cinematic close-up: a vertical red neon sign reading "夜行者客栈" on a rainy neon street, raindrops sliding down the tubes; below it the swordswoman in a black coat glances back, sword-hilt filigree sharp. Shallow depth of field, bokeh neon, crimson vs. cold teal, cinematic realism, crisp legible text, no watermark.
Using the three reference photos (a woman, a man and a corgi), create a cozy family portrait on a sofa in a sunlit living room. Keep each face and the dog’s markings identical to the references. Warm natural light, shallow depth of field, editorial photo style, no text.
Three steps, no software to install.
Write the subject, style, lighting and any text exactly as it should appear, in quotes.
Upload up to 14 photos of people, products or styles to keep consistent, or one photo to edit.
Choose 1K, 2K or 4K and an aspect ratio. Download the result or keep editing it.
Nano Banana 2.1 is billed per image by output resolution — about $0.30 at 1K, $0.45 at 2K and $0.65 at 4K at the base rate (20 credits = $1).
| Example | Rate | Total |
|---|---|---|
| 1 image · 1K | 6 credits | 6 credits |
| 1 image · 2K | 9 credits | 9 credits |
| 1 image · 4K | 13 credits | 13 credits |
Each reference image adds 1 credit (up to 14). Failed generations are refunded automatically.
About Nano Banana 2.1 on SeeBloom.
Nano Banana 2.1 is the nickname for Google’s next-generation Gemini Flash image model. It generates and edits images with accurate text rendering and strong character consistency.
Yes. It renders both Chinese and English text clearly. Put the exact words in quotes in your prompt for the best result.
Up to 14 reference images per generation — useful for keeping several characters or products consistent.
Yes. New accounts get 30 free credits — enough for 3 images at 2K.
Nano Banana 2.1 is great for typography, multi-reference consistency and 4K output. GPT Image 2.5 excels at precise instruction-following edits, masked inpainting and transparent backgrounds.
No. SeeBloom is an independent platform providing access to Nano Banana 2.1 alongside other leading models.
Sign up to get free credits and create your first image in seconds.
Open Nano Banana 2.1 in SeeBloom →
SeeBloom