Nano Banana 2 AI Image Generator

Google’s Nano Banana 2 turns a specific brief into a finished still: the objects you name, the light you ask for, and the frame you pick. Start from text, or bring up to 14 references when you are editing.

01 — Capabilities

A brief it can actually hold

Nano Banana 2 is built for instructions with several parts — objects, materials, light and frame — and for edits that have to stay faithful to what you upload.

  1. Every object you name

    01

    Every object you name

    A cup, a fig, a spoon, a lemon: list them and they stay separate instead of melting into one still-life. Long prompts are read through, up to 12,000 characters.

  2. A person you can specify

    02

    A person you can specify

    Hair, coat, place and lens can all sit in the same sentence. When you are editing, add up to 14 reference images so a face, product or style stays put.

  3. A frame for the shot, not a default

    03

    A frame for the shot, not a default

    Square, story, widescreen or a strip. Ratios run from 1:1 and 16:9 through 21:9, and out to 1:4, 4:1, 1:8 and 8:1, or auto to follow a source image.

  4. Photo, or something you can touch

    04

    Photo, or something you can touch

    Photoreal light and materials, or a built surface like clay. Resolution steps from a 0.5K draft to 1K, 2K and 4K when the still has to hold up large.

02 — Prompt to output

A prompt, then the still

Each still ships with the prompt that made it. Open one in Create and change it.

  • Portrait01 · Portrait

    PromptClose photographic portrait of a young man with curly hair wearing a grey knit sweater, soft window light, he glances aside and smiles warmly, shallow depth of field.

  • Product02 · Product

    PromptA clean white leather sneaker on a pale grey seamless studio backdrop, soft diffused light, three-quarter view, crisp stitching, minimal commercial photograph.

  • Street03 · Street

    PromptPhotographic view down a cobblestone lane in an old European village, climbing roses on stone walls, café tables in morning sun, a few people strolling in the distance.

  • Character04 · Character

    PromptPhotograph of a young woman with shoulder-length brown hair in an olive t-shirt, golden late-afternoon light, she turns her head and looks into the distance, breeze in her hair.

03 — Where it fits

Where a still has to be specific

Work that starts from a brief — a product, a person, a place — rather than a vague mood.

  • Campaign stills

    Campaign stills

    Hero product frames with room for a headline. One clean object, controlled light, empty space where type will sit.

  • Social portraits

    Social portraits

    Vertical frames for a profile, a drop or a cover, with the subject placed so a caption can live in the corner.

  • Concept frames

    Concept frames

    A place that does not exist yet, specific enough to show a client before anyone builds it.

  • Brand spaces

    Brand spaces

    Interiors and sets where materials, light and emptiness matter as much as the object in the room.

04 — Verdict

What Nano Banana 2 is good at

Strengths

  • Holds multi-part prompts, including prompts up to 12,000 characters
  • Edits from as many as 14 reference images
  • Frames from standard ratios out to 1:8 and 8:1, plus auto
  • Output from a 0.5K draft up to 4K

Trade-offs

  • One image per run, so each variation is another generation
  • 0.5K is a draft tier — too soft for anything you will publish
  • 1:8 and 8:1 only work when the subject is designed as a strip

05 — Side by side

Nano Banana 2 vs GPT Image 2.5 vs Seedream

Three image models you can open on ImgToVid, on the numbers that change a job.

Nano Banana 2GPT Image 2.5Seedream
DeveloperGoogleOpenAIByteDance
Top resolution4K4K4K on 5.0 Lite
Reference imagesUp to 14Up to 16Up to 15 on Lite
Images per run11Up to 4 on Lite
Extreme frames1:8 and 8:11:3 and 3:1Standard set
On ImgToVidAvailable nowAvailable nowAvailable now

06 — How it works

Three steps to a Nano Banana 2 still

  1. Step 1

    Pick the version

    Nano Banana 2 for range. Switch to 2.1, 2 Lite or Pro from the family below when the job asks for something else.

  2. Step 2

    Write the still

    Name the subject, the light and the frame. Upload reference images when you are editing an existing picture.

  3. Step 3

    Generate in Create

    Set ratio and resolution, then generate. Your prompt and settings carry straight into Create.

07 — Versions

The Nano Banana versions on ImgToVid

Same prompt box in Create. The limits are what change.

08 — FAQ

Nano Banana 2 questions

What is Nano Banana 2?

Nano Banana 2 is Google’s image model for text-to-image and image editing. On ImgToVid it outputs one image per run, from 0.5K up to 4K, and an edit can take up to 14 reference images.

How is it different from Nano Banana 2.1, Lite and Pro?

Nano Banana 2 covers 0.5K to 4K and extreme ratios including 1:8 and 8:1. Nano Banana 2.1 is 1K to 4K, and those extreme ratios are only available at 1K. Lite is 1K only, without extreme ratios. Pro is 1K to 4K on a standard set of ratios.

Can I edit my own image?

Yes. Upload one or more images and describe the change. An edit accepts up to 14 images, so you can hold a person, a product and a style reference in the same request.

Which frames and resolutions can I use?

Resolutions are 0.5K, 1K, 2K and 4K. Ratios include auto, 1:1, 3:2, 2:3, 4:3, 3:4, 16:9, 9:16, 5:4, 4:5, 21:9, 1:4, 4:1, 1:8 and 8:1.

Can I use the images commercially?

Commercial use depends on your ImgToVid plan, so check Pricing and our Terms before publishing. Either way, make sure your prompts and uploads don’t infringe on anyone else’s rights.

Write the still the way you would brief it

Pick Nano Banana 2, set the frame, and generate in Create. The prompt you write here is the one that runs.

Start with Nano Banana 2