Skip to main content

Nano Banana Edit: AI Image Editor Online

AI Image EditorFrom 5 credits★ 4.2 / 5Strict content filterUpdated October 2026
Generate with Nano Banana Edit

Bottom line

Nano Banana Edit is Google's multi-reference photo editor: up to 10 images per edit, ten aspect ratios, a flat 5 credits per edit with no resolution tiers. Open the Image Edit tool, select Nano Banana Edit, upload your image, and the exact credit cost appears before the edit runs.

Example output generated with Nano Banana Edit

Nano Banana Edit is the editing version of Google's Nano Banana image model. You upload one to ten photos, type the change in plain language ('put the woman from photo 1 in the jacket from photo 2, keep her face and pose'), and it returns the edited image. On VdoBloom it costs a flat 5 credits per edit. The number of reference photos and the aspect ratio you pick do not change that price, and there is no resolution selector to think about.

It is built for people who edit the same face or product again and again: headshots, outfit swaps, a mascot dropped into new scenes, product composites built from a packshot plus a backdrop. VdoBloom's own tools lean on it for exactly that. AI Headshots, AI Fashion Try-On, AI Product Photography, Real Estate Photography, Art Studio, the GIF generator and Keyframe Morph's end frame all run on this model.

New accounts get 10 free credits, which covers two edits. After that, a one-time 300-credit pack is $19.50 and never expires (60 edits), or the Lite plan is $15 a month for 300 credits. Below: the spec table, worked costs, how it compares with Nano Banana 2.1, Nano Banana Pro and Nano Banana 2, where it runs, prompting notes and the limits to know before you spend credits.

Nano Banana Edit specs at a glance

Everything below comes from VdoBloom's model configuration and credit table. The one thing people expect and do not find is a resolution control: Nano Banana Edit has none. If you need to choose 1K, 2K or 4K, use Nano Banana 2.1 or Nano Banana Pro.

SpecNano Banana Edit on VdoBloom
MakerGoogle (Nano Banana family), run through VdoBloom's KIE integration
ModeImage editing only: at least one uploaded photo is required
Reference images1 to 10 per edit
Aspect ratios10: 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9
Resolution controlNone (no 1K/2K/4K selector)
Output formatPNG or JPEG
Price5 credits per edit, flat (key google/nano-banana-edit)
Free credits cover2 edits on the 10 signup credits
Content filterSTRICT label: Google's provider-side filters apply in full
Where it runsImage Edit, AI Headshots, AI Fashion Try-On, AI Product Photography, Real Estate Photography, Art Studio, GIF generator, Keyframe Morph, Pose Variations, Studio

What Nano Banana Edit costs: worked examples

One balance of credits covers every model on VdoBloom. The dollar column uses the Lite plan rate of $15 for 300 credits ($0.05 a credit). A one-time 300-credit pack is $19.50 and those credits never expire; monthly plan credits reset each billing cycle. A failed edit is refunded automatically.

JobCreditsAt the Lite rate
1 edit (any ratio, 1 to 10 references)5$0.25
2 edits (the free signup credits)10free
10 edits50$2.50
20 edits100$5.00
60 edits (one 300-credit pack)300$15.00 on Lite, $19.50 one-time

Read more: Nano Banana Edit pricing explained · Credit packs and plans

Nano Banana Edit vs Nano Banana 2.1 vs Nano Banana Pro vs Nano Banana 2

All four sit in the same Image Edit dropdown, so switching is one click and your prompt stays in the box. The plain summary: Nano Banana Edit is the flat-price, no-settings option. Nano Banana 2.1 costs less than it at 1K and 2K (3 and 4 credits) and takes the same 10 references, but asks you to pick a resolution. Nano Banana Pro adds 1K/2K/4K control at 18 to 24 credits with up to 8 references. Nano Banana 2 is a flat 15 credits with 10 references and the extreme banner ratios. VdoBloom does not publish a quality ranking between them; run the same edit through two and keep the one you prefer.

ModelMax referencesResolutionAspect ratiosCredits per edit
Nano Banana Edit10No selector105 flat
Nano Banana 2.1101K / 2K / 4K15 incl. auto, 8:1, 1:83 / 4 / 6
Nano Banana Pro81K / 2K / 4K1018 / 18 / 24
Nano Banana 2101K / 2K / 4K15 incl. auto, 8:1, 1:815 flat
Nano Banana 2 Lite10No selector15 incl. auto8 flat

Read more: Nano Banana 2.1 model page · Nano Banana Pro model page · Nano Banana (text to image) model page · Nano Banana vs Nano Banana Edit

Nano Banana vs Nano Banana Edit: which page do you need?

They are two halves of one family. Nano Banana (google/nano-banana) makes an image from text alone and costs 6 credits. Nano Banana Edit (google/nano-banana-edit) needs at least one uploaded photo and costs 5. If you start with a photo, you want Edit. If you start with a blank canvas, you want the generator, and you can feed its result into Edit afterwards for the next change. Both share the same ten aspect ratios and the same STRICT content label.

Where to run Nano Banana Edit on VdoBloom

The model is the same everywhere; what changes is who writes the prompt. In Image Edit you write it. In the purpose-built tools below, VdoBloom writes a tested prompt for you and you only supply photos and pick options. Each tool shows its own credit total before you generate, because some return several images per run.

  • Image Edit: pick Nano Banana Edit in the dropdown, upload up to 10 photos, choose one of ten aspect ratios and describe the change. 5 credits per edit.
  • AI Headshots: one clear photo in, styled headshots out. It runs Nano Banana Edit at low edit strength so the face stays locked and only background, outfit and lighting change.
  • AI Fashion Try-On (Beta): a person photo plus an outfit photo, rendered as the person wearing that garment. Beta because it uses this general multi-image edit model rather than a dedicated try-on model.
  • AI Product Photography and Real Estate Photography: a packshot or a room photo restyled into a finished scene.
  • Art Studio, the GIF generator and Pose Variations list it as an edit model; Keyframe Morph uses it to build the end frame of a morph video; the Manga generator uses it when you upload a character reference.

Read more: AI Headshot Generator from one photo · AI Fashion Try-On · AI Product Photography · AI Image Editor: all edit models in one place · Step-by-step: how to use Nano Banana Edit

Prompting Nano Banana Edit

Say what changes and, just as clearly, what must not. 'Replace the background with a sunlit cafe; keep her face, hair, pose and the white shirt exactly as they are' gives the model a fence. A prompt that only describes the new scene leaves it free to re-dress or re-light the person as well.

With several references, name them by order and role: 'Photo 1 is the person, photo 2 is the jacket, photo 3 is the location.' Use one photo per thing you want copied, and pick the clearest front-facing shot of a face rather than three similar selfies. Choose the aspect ratio before you generate; the output follows the ratio you pick, so a 4:5 feed post should be set to 4:5 rather than cropped later. Because every edit is 5 credits, the cheap fix for a near-miss is a second run with a tighter fence, not a longer description of the scene.

How many reference photos to use, and which

Ten is the limit, and the price is the same 5 credits whether you send one photo or ten, so extra references cost no credits, only clarity. Each photo you add is something the model has to reconcile with the others. A good rule is one photo per thing you want carried into the result.

JobReferences that workWhat to say in the prompt
New background, same person1: the person'Keep face, hair, pose and clothing exactly; replace only the background with ...'
Outfit swap2: the person, the garment'Photo 1 is the person, photo 2 is the jacket; dress the person in the jacket, keep face and pose'
Product in a scene2 to 3: packshot, backdrop, optional lighting reference'Place the bottle from photo 1 on the counter from photo 2; keep the label legible'
Character series1 to 2 clear shots of the character'Same character as photo 1, now sitting at a cafe table, same outfit and colours'
Group composite1 per person, plus the locationName each person by photo number and give each a position in the frame

Why VdoBloom's people tools are built on it

Several of VdoBloom's ready-made tools could run on any edit model, and they were built on this one. AI Headshots sends your photo with a low edit strength, so the model changes the background, clothes and light while the face stays the face. AI Fashion Try-On sends two images, the person and the garment, and asks for one combined result; VdoBloom had already validated the model for the same two-image combine in its UGC ad flow. AI Product Photography and Real Estate Photography use it to restyle one photo into finished scenes. Keyframe Morph uses it to make the end frame of a morph from your start frame, and the Manga generator switches to it when you upload a reference for a character.

The reason is the same each time: a flat 5-credit price that does not grow with references, ten reference slots, and an edit that tends to keep the subject recognisable. If you like what one of those tools produces, you can get the same model with your own prompt in Image Edit.

Limits and honest caveats

There is no resolution control and no 4K option on this model. For a print-size deliverable, use Nano Banana 2.1 (6 credits at 4K) or Nano Banana Pro (24 credits at 4K). Ten references is the ceiling, the same as on Nano Banana 2 and Nano Banana 2.1. It cannot generate from text alone, so an edit with no uploaded photo is refused.

The STRICT label means Google's provider filters apply to both your prompt and your uploads. Swimwear, suggestive poses and some images of real public figures are more likely to be declined here than on Seedream 4.5 Edit, which carries the looser HOT label. Instruction drift is the common complaint for this whole family: an edit that changes more than you asked, or a likeness that softens across several references. The mitigation is the 'keep exactly' clause above and fewer, clearer references. Like every Nano Banana release it is proprietary: VdoBloom runs it for you, there are no downloadable weights, and if the provider fails a job the credits come back automatically.

Read more: Nano Banana blocking fashion or swimwear images? Alternatives

What creators use Nano Banana Edit for

  • Identity-preserving people edits: new background, outfit or lighting while the face stays the face from the reference photo.
  • Outfit and product composites built from several photos at once: a person, a garment, a packshot and a backdrop in one instruction, up to 10 images.
  • Feed-sized edits in 4:5 or 5:4, plus 9:16 for Stories and 21:9 for banners, without cropping afterwards.
  • Batch work where a predictable price matters: every edit is 5 credits whether it uses one reference or ten.
  • Mascot and character continuity: the same character placed into a series of new scenes for a campaign or a comic page.

How to use Nano Banana Edit on VdoBloom

  1. 1Open the Image Edit tool and select Nano Banana Edit.
  2. 2Upload the image(s) to edit and describe the change you want.
  3. 3Review the credit cost and generate. Results appear in your library.

Frequently asked questions

How much does Nano Banana Edit cost on VdoBloom?

A flat 5 credits per edit. Using one reference photo or ten, and picking any of the ten aspect ratios, costs the same. At the Lite plan rate of $0.05 a credit that is $0.25 an edit; a one-time 300-credit pack ($19.50) covers 60 edits.

Can I use Nano Banana Edit for free?

Yes, for two edits. Every new account gets 10 free credits and each edit is 5. More credits come from referrals and challenges, or from a one-time pack whose credits never expire.

How many images can Nano Banana Edit use at once?

Up to 10 per edit. VdoBloom rejects an 11th. Use them for distinct roles (a person, a garment, a product, a location) rather than several near-identical shots of the same thing.

Does Nano Banana Edit keep the same face?

Keeping identity is what VdoBloom uses it for: AI Headshots runs it at low edit strength so only background, outfit and lighting change. Results still vary by photo, so say 'keep the face and hair exactly as in photo 1' and use a clear, front-facing reference.

What resolution does Nano Banana Edit output?

There is no resolution selector on this model. You choose the aspect ratio and the output format (PNG or JPEG). If you need to pick 1K, 2K or 4K, Nano Banana 2.1 (3, 4 or 6 credits) and Nano Banana Pro (18, 18 or 24) both offer it.

Which aspect ratios does Nano Banana Edit support?

Ten: 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9 and 21:9. There is no auto option and no 8:1 or 1:8 banner strip; those are on Nano Banana 2 and 2.1.

Nano Banana Edit vs Nano Banana 2.1: which should I use?

Nano Banana 2.1 costs less at 1K and 2K (3 and 4 credits against 5), takes the same 10 references and adds 1K/2K/4K plus banner ratios. Nano Banana Edit is the flat 5-credit option with no settings, and it is the model VdoBloom's headshot, try-on and product-photo tools are built on. Run your edit on both; the price difference is one or two credits.

Why did Nano Banana Edit refuse my photo?

It carries the STRICT content label, so Google's filters check both the prompt and the uploaded images. Revealing outfits, suggestive poses and some real-person edits are declined more often than on models labelled HOT, such as Seedream 4.5 Edit. A refused or failed job is refunded automatically.

Who makes Nano Banana, and is it the same as Gemini?

Google. Nano Banana began as the nickname for the image model in Google's Gemini family and stuck as its public name. VdoBloom runs it through its own provider integration, so you do not need a Google account, a Gemini subscription or an API key.

Is Nano Banana Edit open source?

No. It is a proprietary Google model with no downloadable weights. You use it through a hosted service such as VdoBloom, which handles the provider relationship and bills in credits.

Nano Banana Edit guides, prices and comparisons

This page is the Nano Banana Edit hub. Each guide below goes deeper on one question and links back here.

Related models