Video Generation API is now live!
Nano Banana ProImage API

Nano Banana Pro APIfor developers.

Edit, relight, restyle, localize, and compose images with Google's quality-first Gemini 3 Pro Image model.One workflow for multi-image composition, text rendering, identity preservation, and production-grade edits up to 4K.

see more

Modes

Edit, compose, transform

Built primarily for image editing and multi-image composition through natural-language instructions

Resolution

1K, 2K, 4K

fal.ai documents 1K / 2K / 4K options, with 4K billed at a premium

Aspect ratios

Auto to 21:9

Supports wide, square, portrait, and story-style layouts depending on provider settings

Multi-image input

Up to 14 images

Useful for composition, style reference, product context, and controlled blends

Identity consistency

Up to 5 people

Designed to preserve resemblance more reliably across edits and compositions

Pricing

From $0.14 to $0.15 / edit

WaveSpeed lists $0.14 for 1K/2K edits; fal.ai lists $0.15 per image with 4K at 2x cost

Available Endpoints

Start building with the Nano Banana Pro API

Multiple endpoints for text-to-video, image-to-video, fast preview flows, and async job retrieval. This section is laid out more like a product catalog than raw docs so users can scan what to use first.

NewQuality-first

Endpoint

fal.ai Edit API

fal-ai/nano-banana-pro/edit

Image editingfal.aiCommercial use1K / 2K / 4K

Edit one or more images with prompt-driven semantic changes, composition-aware transforms, and provider-managed output handling.

Best for: Use this when you want a hosted editing API with direct fal.ai tooling, commercial terms, and quality-first Gemini 3 Pro Image behavior.

NewREST

Endpoint

WaveSpeed Edit API

google/nano-banana-pro/edit

WaveSpeedSync modeBase64 outputREST API

WaveSpeed exposes Nano Banana Pro Edit with prompt, images, aspect ratio, resolution, output format, sync mode, and optional base64 output controls.

Best for: A good fit when you want a ready-to-use REST path with provider-specific controls like synchronous mode or base64 responses.

NewRoute

Endpoint

OpenRouter Route

google/gemini-3-pro-image-preview

OpenRouterNormalized APIImage + textProvider routing

OpenRouter normalizes access to Google's Gemini 3 Pro Image Preview route and returns image outputs through its chat-style API format.

Best for: Best when you want OpenRouter's provider routing, pricing abstraction, or one API surface across many image and language models.

New

Endpoint

Reference Composition

images[] + prompt

Multi-imageCompositionReference-guidedContext-aware

The model accepts multiple image inputs so you can blend subject, style, layout, and context into one edited result.

Best for: Use this when product teams need in-context compositing, virtual staging, style transfer, or preserve-and-change workflows.

Get started today

Ready to integrate Nano Banana Pro?

Try the API directly in the console, or reach out to the team for onboarding, pricing, and enterprise setup.

API Documentation

How to get access to Nano Banana Pro API

Provider integrations differ slightly, but the core workflow is consistent: send image URLs plus a prompt, choose resolution and aspect ratio, then read back the finished edited image output.

Selected endpoint

fal-ai/nano-banana-pro/edit

Start with one or more reference images and a precise edit instruction. Be explicit about what must change and what must stay intact.

Use this when you want a hosted editing API with direct fal.ai tooling, commercial terms, and quality-first Gemini 3 Pro Image behavior.

import { fal } from "@fal-ai/client";

fal.config({
  credentials: process.env.FAL_KEY
});

const result = await fal.subscribe("fal-ai/nano-banana-pro/edit", {
  input: {
    prompt: "Place the sofa from image 2 into image 1 and blend the shadows, perspective, and room lighting naturally. Keep the wall texture and floor reflection realistic.",
    image_urls: [
      "https://example.com/room.jpg",
      "https://example.com/sofa.png"
    ],
    aspect_ratio: "4:3",
    resolution: "2K",
    output_format: "png"
  },
  logs: true
});

console.log(result.data.images[0].url);

Async flow

  1. 1

    Choose the provider path you want to integrate first: fal.ai, WaveSpeed, or OpenRouter.

  2. 2

    Send your prompt together with one or more source image URLs, then set resolution, aspect ratio, and output format based on your delivery target.

  3. 3

    If you are using a queue-aware or async workflow, store the request id and poll or wait for the result until the image is ready.

  4. 4

    Read the output URL or image payload, then move the finished asset into your own review, storage, localization, or publishing flow.

What Makes It Different

What makes the Nano Banana Pro API different

This section is laid out to read more like a product narrative than a feature list. Each row shows a capability, why it matters, and what that looks like in a real workflow.

Preview

Semantic editing without masks

That makes instructions like changing materials, relighting a scene, moving one product into another environment, or swapping outfits much easier to express in plain language.

Capability

Semantic editing without masks

Nano Banana Pro is built to understand what your prompt means in the context of the scene instead of forcing manual selection-heavy workflows.

That makes instructions like changing materials, relighting a scene, moving one product into another environment, or swapping outfits much easier to express in plain language.

Example scenario

A creative team says 'keep the model, keep the pose, replace the set with a premium studio stage, and make the dress velvet' without building a manual mask stack.

Capability

Multi-image composition

fal.ai documents support for up to 14 input images, which is useful when you need one result informed by several references at once.

You can blend subject images, style frames, room references, product renders, and target aesthetics into one more controlled output.

Example scenario

A furniture workflow combines the base room, a sofa cutout, and a lighting reference to create one realistic in-context product visual.

Preview

Multi-image composition

You can blend subject images, style frames, room references, product renders, and target aesthetics into one more controlled output.

Preview

Text rendering and multilingual edits

That makes the model much more practical for packaging, banners, UI mockups, posters, educational visuals, and localized campaign assets.

Capability

Text rendering and multilingual edits

WaveSpeed and OpenRouter both describe stronger text rendering, longer passages, and multilingual on-image text handling as key differentiators of the Gemini 3 Pro Image stack.

That makes the model much more practical for packaging, banners, UI mockups, posters, educational visuals, and localized campaign assets.

Example scenario

A growth team updates a global ad set by preserving the scene while translating the in-image headline into Portuguese or Japanese.

Capability

Quality-first 4K workflows

The model is positioned as quality-first rather than speed-first, with 1K, 2K, and 4K outputs depending on the provider.

It is better suited for premium campaign edits, product stills, and client-facing design work where fidelity matters more than low-latency iteration.

Example scenario

A brand team uses 1K or 2K for review rounds, then reruns the approved edit at 4K for final campaign delivery.

Preview

Quality-first 4K workflows

It is better suited for premium campaign edits, product stills, and client-facing design work where fidelity matters more than low-latency iteration.

Unified API Platform

Two API tiers for different use cases

Pick the right balance of quality, speed, and cost for your workflow. The section stays data-driven, but the presentation is closer to a clean product comparison table.

Feature
1K
2KRecommended
4K
Best forFastest review loopsBalanced quality and costHighest fidelity visual delivery
SpeedFastestModerateSlowest
QualityBest for internal review and prompt tuningGood for many production-facing editsHighest detail and cleanest edges
CostLowest published costMid-tierHighest
Recommended useEarly edit direction, rough comp validation, and frequent iterationMarketing visuals, product edits, and most customer-facing image workflowsPremium product imagery, client review, polished campaign exports, and print-like compositions
API endpointsfal-ai/nano-banana-pro/edit, google/nano-banana-pro/editfal-ai/nano-banana-pro/edit, google/nano-banana-pro/editfal-ai/nano-banana-pro/edit, google/nano-banana-pro/edit

Use Cases

Industries using the Nano Banana Pro API

This section keeps the same reusable data model, but the presentation is closer to a grid of industry cards instead of long narrative boxes.

Merchandising teams and creative ops

Product photo refinement

Adjust materials, relight scenes, swap props, and refine premium product stills without rebuilding the shot from scratch.

Nano Banana Pro is especially useful when product fidelity, reflections, and text on labels need to survive the edit.

Furniture, fashion, and lifestyle brands

In-context composition

Place one object or product into another scene and let the model align perspective, shadows, and visual coherence.

This reduces the need for manual compositing or separate 3D staging in many everyday marketing workflows.

Growth teams and international marketers

Localization and text edits

Translate or replace in-image copy while preserving the rest of the creative direction and layout.

Text rendering and multilingual image editing are a practical differentiator for real campaign operations.

Agencies, creator tools, and social teams

Character-consistent ad iteration

Maintain resemblance for people across multiple edit passes while changing styling, backgrounds, lighting, or scene context.

That makes the model more usable for campaign systems that need repeated edits instead of one-off generations.

Design teams and art direction workflows

Concept art and stylization

Restyle scenes into cinematic, illustrated, game-like, or culture-specific visual treatments from natural-language prompts.

The model supports both photoreal refinement and stronger style shifts without abandoning the original composition intent.

Studios and client-facing production teams

High-fidelity review output

Move from 1K and 2K validation to 4K-style premium exports when the asset is approved.

A quality-first model is more useful when the final deliverable must hold up beyond internal prototyping.

Examples

Nano Banana Pro API examples

This section supports future video or image previews, but it also renders cleanly from prompt-only data. That keeps the template reusable even when a page launches before media assets are ready.

Demo

Glass hamburger transformation

Material transformation

A classic preserve-and-change edit: keep the layout, change the core material, and let the model handle lighting coherence.

Make the hamburger made of glass, keep the composition intact, preserve the soft studio lighting, and render the material as translucent with realistic refraction.

material editproductsemantic edit
Demo

Interior product composition

Multi-image composition

This is one of the strongest use cases for Nano Banana Pro: combining references into one believable scene without heavy manual compositing.

Place the sofa from image 2 into image 1 and blend the light, shadows, color temperature, and perspective so the scene looks naturally photographed together.

multi-imagecompositioninterior
Demo

Nighttime scene conversion

Relighting

Use prompts like this when your goal is environmental transformation without changing the scene layout itself.

Turn this daytime street scene into nighttime. Preserve perspective, architecture, signage, and reflections while adding realistic street lighting and a cooler cinematic mood.

night moderelightingscene edit
Demo

Localized ad copy update

Multilingual text edit

A practical fit for international marketing workflows where the creative should stay stable while the in-image message changes by market.

Translate the text in the image into Brazilian Portuguese while preserving the composition, hierarchy, visual balance, and overall campaign style.

localizationtext renderingcampaign

How To Use This API

How to use Nano Banana Pro API

This quick-start walkthrough is written to rank for integration-style searches while staying concise enough for busy developers and operators.

  1. 1

    Choose your provider path

    Start with fal.ai, WaveSpeed, or OpenRouter depending on whether you want a direct edit API, provider-specific REST controls, or a normalized multi-model routing layer.

  2. 2

    Prepare one or more image URLs

    Use clean source images and gather any additional reference visuals you want the model to blend, preserve, or learn style from.

  3. 3

    Write preserve-and-change instructions

    Be explicit about what must change and what must stay untouched. Nano Banana Pro performs best when the prompt separates those two ideas clearly.

  4. 4

    Pick resolution and aspect ratio

    Use 1K or 2K for iteration, then rerun the approved edit at higher resolution when you need premium output or client delivery quality.

  5. 5

    Review output and iterate

    Once you have the first result, tighten the prompt toward finer material, lighting, text, or composition changes instead of restarting from a vague instruction.

FAQ

Frequently asked questions about Nano Banana Pro API

FAQs stay compact and skimmable here. The content is still data-driven for SEO, but the layout is cleaner and less visually heavy.

What is Nano Banana Pro API?

It is the provider-hosted API surface for Google's Gemini 3 Pro Image model, optimized for image editing, multi-image composition, text rendering, and quality-first visual transformations.

Is Nano Banana Pro only for editing?

The strongest provider pages emphasize editing and composition, but OpenRouter also positions Gemini 3 Pro Image Preview as a broader image-generation and image-editing model for professional visual workflows.

What makes it different from the original Nano Banana?

The cited provider pages position Nano Banana Pro as the higher-quality Gemini 3 Pro Image tier, trading pure speed for stronger reasoning, sharper text rendering, richer composition control, and better character consistency.

How many images can I provide?

fal.ai documents support for combining up to 14 images in a single composition workflow.

What resolutions are available?

Provider documentation references 1K, 2K, and 4K-style output tiers, with higher resolutions carrying a higher unit cost.

Does it support commercial use?

fal.ai explicitly marks Nano Banana Pro Edit as commercial use permitted. You should still review each provider's own terms and any downstream publishing requirements.

Can it preserve people or identity across edits?

The provider documentation highlights resemblance and identity preservation for up to 5 people, which makes it more useful for character-consistent marketing and editing workflows.

Which provider should I start with?

Use fal.ai if you want a direct model API with clear edit semantics, WaveSpeed if you want REST controls like synchronous mode and base64 output, and OpenRouter if you want normalized multi-provider routing.

Model Directory

Browse the full model market before you choose your route.

Use the `/models` catalog to scan providers, modalities, reasoning support, context windows, and pricing metadata from a local OpenRouter snapshot. It is the fastest way to compare what exists before you decide which models should be prioritized on ImaRouter.

Get Started

Validate Nano Banana Pro on the provider you trust, then productionize the workflow that sticks

Use WaveSpeed for quick REST-style validation, fal.ai for direct edit integration, or OpenRouter for normalized routed access. Once the preserve-and-change prompt pattern is stable, you can move the exact workflow into production.