Nano Banana

Text to ImageNatural-Language EditingMulti-Image Input

Nano Banana on APIXO maps to Google’s stable Gemini 2.5 Flash Image model. Create new 1K-class images from prompts or transform existing visuals with natural-language instructions, multi-image context, flexible framing, and JPEG or PNG delivery—without mixing in Nano Banana 2 or Nano Banana Pro capabilities.

Upstream Model

Gemini 2.5 Flash Image

APIXO Price

$0.03 / Image

Image Input

Up to 5 Images

Output

1K-Class · JPEG or PNG

Released

August 26, 2025

Loading workspace...

Model Strengths

Direct creative control through language and visual context

Prompt-to-Image Creation

Turn detailed text instructions into complete 1024px visuals spanning photographic, illustrative, and designed compositions.

Natural-Language Editing

Request targeted changes to objects, backgrounds, colors, lighting, or composition without supplying a separate mask.

Multi-Image Composition

Combine visual context from several source images into a newly generated scene or coordinated asset.

Subject Continuity

Carry recognizable subjects and recurring visual traits into revised scenes, while allowing for normal generative variation.

World-Aware Prompts

Use Gemini’s multimodal understanding and world knowledge to interpret contextual instructions and relationships within a scene.

Style Transformation

Rework the presentation of an image through descriptive art direction while retaining the source as visual context.

APIXO Workflows

Start from a prompt or reshape existing imagery

Text to Image

Submit a required prompt and choose an APIXO aspect ratio and output format. This workflow creates a new image without reference-image input.

Create from text

Image to Image

Provide one to five image URLs with a required instruction. The images act as editable content or general visual context rather than dedicated style or identity-reference controls.

Edit or compose

Production Specifications

The controls exposed by APIXO

A focused interface for prompt-led creation and independent image-editing tasks.

1K-class

Output Resolution

1–5 Images

Image-to-Image Input

11 Options

Aspect Ratios Including Auto

JPEG · PNG

Output Formats

Async / Callback

Delivery

Prompt-Based

Editing Guidance

Production Uses

Move from source material to usable campaign imagery

Nano Banana product photography recontextualized into a styled interior

Commerce

Recontextualize product photography

Place a supplied product or subject into fresh environments, lighting setups, and campaign compositions for concept development. Review geometry and branding before publication.

Nano Banana creative reframed for an ultrawide placement

Marketing

Adapt creative for new placements

Reframe an established visual for square, portrait, landscape, or ultrawide placements while directing background and compositional changes in natural language.

Nano Banana recurring character explored across successive scenes

Story Development

Explore recurring characters and scenes

Build storyboards, editorial concepts, or social content around recognizable subjects across successive independent generations.

Nano Banana references combined into a coordinated fashion concept

Creative Direction

Combine references into one concept

Use several images as visual context for mood exploration, composite concepts, merchandising ideas, or presentation-ready creative directions.

Before You Build

Operational details and practical boundaries

Implementation Notes

APIXO’s Nano Banana route maps to Google’s stable Gemini 2.5 Flash Image model, not Nano Banana 2 or Nano Banana Pro.

APIXO accepts as many as five image URLs for image-to-image requests. Google states that the upstream model works best with up to three input images.

Reference images provide editable content or general visual context; APIXO does not expose separate subject-, character-, or style-reference fields.

Each APIXO request is independent. The public route does not expose Gemini’s conversational, stateful multi-turn editing workflow.

APIXO documents typical processing around 10–15 seconds, but actual completion time can vary with service demand.

Google applies SynthID watermarking to generated images. APIXO does not publicly expose mask, seed, transparent-output, or batch-count controls on this route.

Frequently Asked Questions

It is Google’s Gemini 2.5 Flash Image, identified upstream as gemini-2.5-flash-image and released on August 26, 2025. It is the original Nano Banana model, separate from Nano Banana 2 and Nano Banana Pro.

Yes. APIXO’s image-to-image workflow accepts one to five image URLs plus a required natural-language instruction. No mask parameter is publicly exposed.

APIXO publicly lists a flat price of $0.03 per generated image for this model. No separate workflow or resolution tier is documented.

APIXO exposes 1:1, 9:16, 16:9, 4:3, 3:4, 3:2, 2:3, 5:4, 4:5, 21:9, and automatic aspect-ratio selection.

Review hands, anatomy, embedded text, logos, exact facial identity, product geometry, and reference fidelity. Generative edits are not deterministic, and repeated requests can produce different results.