Nano Banana Pro API in 2026: Official Gemini Route, Pricing, Code, and Provider Checks

Nano Banana Pro API is the Gemini API route for Google Gemini 3 Pro Image. Use gemini-3-pro-image for the current official Pro model, expect paid image output, and choose the contract owner before you copy code.

Official Google Standard image output is $0.134 for 1K/2K and $0.24 for 4K; Google Batch/Flex halves those image prices to $0.067 and $0.12 when asynchronous delivery is acceptable. laozhang.ai can be useful as a provider gateway, but its public $0.09/image row is provider pricing, not official Google pricing, and the console price must prevail before production.

The current route decision looks like this:

These API facts were checked on July 1, 2026 against Google AI image-generation, pricing, API-key, billing, and rate-limit docs, plus public laozhang.ai Nano Banana Pro image docs. Prices, model names, provider aliases, quota tiers, and billing behavior can change; recheck the official docs or provider console before launch.

What Nano Banana Pro API Means Today

Nano Banana Pro API is the reader-facing name many developers use, but the official Google mapping is Gemini 3 Pro Image through the Gemini API. The current official model ID is gemini-3-pro-image. Older strings such as gemini-3-pro-image-preview should be treated as legacy examples or provider aliases unless the current Google docs or your chosen provider console explicitly require them.

This distinction matters because a working request has three owners:

• Model owner: the actual image model you call, such as gemini-3-pro-image.
• Contract owner: Google Standard, Google Batch/Flex, Vertex/Gemini API, or a provider gateway.
• Operational owner: the project, key, quota, logs, and support surface that will explain failures.

If you skip the contract decision, code examples become misleading. The same prompt can be cheap enough on Nano Banana 2, expensive but necessary on Nano Banana Pro, slower but cheaper through Batch/Flex, or operationally acceptable through a gateway only after proof.

Choose The Model Before You Optimize The Route

Nano Banana Pro is the override model, not the default model for every image job. Use it when 4K output, complex layout, typography fidelity, or professional-grade detail is the reason the image exists. Use a cheaper Nano Banana model when the job is exploration, draft imagery, simple product shots, or high-volume iteration.

The cost-saving move is often model substitution, not gateway shopping. If you are not sure whether Pro is necessary, compare the output against Nano Banana Pro vs Nano Banana 2 before building a Pro-only pipeline.

Official Google Pricing: Standard Versus Batch/Flex

Google pricing should stay in its own table. Do not mix provider rows into official rows.

The important reading is not only "Batch/Flex is cheaper." It is why it is cheaper: you are changing the official processing lane, not discovering a hidden Nano Banana Pro coupon. Keep these two rules close to every estimate:

• No official free Pro image-output tier: the pricing page lists no free tier row for gemini-3-pro-image image output.
• Grounding can add cost: if a request uses Google Search or image grounding, search-query charges can apply separately from image output.

If your only question is the Pro price table, the deeper cost breakdown belongs in Nano Banana Pro pricing or Gemini 3 Pro Image API pricing. Keep pricing tied to route choice and implementation readiness, not to isolated unit rows.

Key, Billing, And Quota Are Project Boundaries

Create or view Gemini API keys in Google AI Studio, but do not confuse the credential with the billing and quota owner. Google's current key docs say new AI Studio keys are auth keys; unrestricted standard keys are rejected from June 19, 2026, and standard keys are rejected in September 2026. That migration belongs in the setup plan, not in a post-launch incident.

Use this pre-code checklist:

If you are specifically blocked by quota, use Increase Nano Banana Pro API quota instead of rotating keys. Key rotation is a security operation; it is not a scaling plan.

First Official Request With The Interactions API

Google's current image-generation docs state that the Interactions API is generally available and recommended for access to the latest features and models. The examples below use the current gemini-3-pro-image model ID. Keep GEMINI_API_KEY in an environment variable, not in source code.

Python

python
from google import genai
import base64

client = genai.Client()

interaction = client.interactions.create(
    model="gemini-3-pro-image",
    input="Create a premium product hero image for a black mechanical keyboard on a clean studio background.",
    response_format={
        "type": "image",
        "aspect_ratio": "16:9",
        "image_size": "2K",
    },
)

if interaction.output_image:
    with open("nano-banana-pro-output.png", "wb") as f:
        f.write(base64.b64decode(interaction.output_image.data))

JavaScript

javascript
import { GoogleGenAI } from "@google/genai";
import * as fs from "node:fs";

const ai = new GoogleGenAI({ apiKey: process.env.GEMINI_API_KEY });

const interaction = await ai.interactions.create({
  model: "gemini-3-pro-image",
  input:
    "Create a premium product hero image for a black mechanical keyboard on a clean studio background.",
  response_format: {
    type: "image",
    aspect_ratio: "16:9",
    image_size: "2K",
  },
});

if (interaction.output_image) {
  fs.writeFileSync(
    "nano-banana-pro-output.png",
    Buffer.from(interaction.output_image.data, "base64"),
  );
}

REST

bash
curl -s -X POST \
  "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-3-pro-image",
    "input": [
      {
        "type": "text",
        "text": "Create a premium product hero image for a black mechanical keyboard on a clean studio background."
      }
    ],
    "response_format": {
      "type": "image",
      "aspect_ratio": "16:9",
      "image_size": "2K"
    }
  }'

For existing codebases still using generateContent, migrate deliberately rather than blindly. The official image docs still expose a generateContent toggle, but new work should use Interactions first unless your SDK, provider, or framework has a clear reason to stay on the older request shape.

Two implementation details save debugging time:

• For image-size labels, keep the K uppercase in values such as 1K, 2K, and 4K.
• For multi-part or interleaved outputs, inspect the response steps/parts instead of assuming one convenience property always captures every generated image.

When A Provider Gateway Such As laozhang.ai Makes Sense

A provider gateway can be a practical route when you want OpenAI-compatible billing, a different payment flow, gateway logs, or a fallback path that is easier for your team to operate. laozhang.ai belongs in the provider lane when the job is API access, cost, quota, and implementation routing.

That lane has a hard boundary: provider pricing is not official Google pricing. Public laozhang.ai docs checked on July 1, 2026 list Nano Banana Pro text-to-image at $0.09/image and image editing at $0.09/edit, and they still expose provider alias language around gemini-3-pro-image-preview. Treat those as provider implementation details. Before production, the provider console, model list, call log, and billable event rule are the source of truth.

Run this proof checklist before routing traffic:

The safe provider recommendation is conditional: use laozhang.ai only when its current console and logs satisfy your price, model, billing, data, and failure requirements. Do not publish or plan around provider claims such as unlimited access, no speed limit, no account risk, no failed-call charge, or guaranteed uptime unless your current contract and logs prove them.

Troubleshooting The First Request

Most first-request failures belong to one of six branches. Keep the branch stable long enough to debug; changing model, key, provider, and payload at the same time destroys the signal.

For broader Nano Banana failure branches, use Nano Banana Pro error codes or Gemini API error troubleshooting. For rate-limit specifics, use Gemini API rate limits guide.

Production Checklist Before Launch

Do not ship a Nano Banana Pro API path until the launch checklist has an owner:

• Model owner: gemini-3-pro-image is required, or a cheaper Nano Banana model passes quality.
• Price owner: Google Standard, Google Batch/Flex, or provider console row is recorded.
• Credential owner: auth key, key restrictions, and rotation policy are clear.
• Quota owner: live project/model limits and tier requirements are checked.
• Cost owner: resolution, retries, grounding, failed calls, and provider billable events are logged.
• Data owner: provider or Google terms fit the image inputs you send.
• Fallback owner: know whether to reduce resolution, move to Batch/Flex, switch to Nano Banana 2, or change provider.
• Rollback owner: keep a lower-cost model or a disabled feature flag ready before traffic rises.

The practical rule is simple: choose the contract owner, prove the price and quota, then copy code. That order keeps Nano Banana Pro API from becoming an expensive debugging exercise.

FAQ

What model ID should I use?

Use gemini-3-pro-image for the current official Google Pro image model. Treat gemini-3-pro-image-preview as a legacy or provider alias unless the active Google docs or your provider console require it.

Is the Pro image API free?

No official free Pro image-output tier is listed for gemini-3-pro-image in the Google pricing page checked on July 1, 2026. AI Studio can be free to open and use as a product surface, but paid API-key usage is still billable.

What are the official Google prices?

Google Standard image output is $0.134 for 1K/2K and $0.24 for 4K. Google Batch/Flex lists $0.067 for 1K/2K and $0.12 for 4K. Recheck the pricing page before production because image rows are volatile.

Is laozhang.ai official Google pricing?

No. laozhang.ai is a provider gateway lane. Its public docs checked on July 1, 2026 list $0.09/image, but that is provider pricing, not official Google pricing. The provider console and call logs must be verified before production.

Do more API keys increase Nano Banana Pro quota?

No. Google rate-limit docs state that limits are applied per project, not per API key. More keys can help with credential rotation or separation, but they do not create more project quota.

Should I use Batch/Flex?

Use Batch/Flex when you need Google's first-party contract and can wait for asynchronous or flexible processing. Do not use it when the product needs immediate interactive output.

Which source should I recheck later?

Recheck Google's image-generation docs, pricing page, API-key docs, billing page, and rate-limit docs. For provider routes, recheck the provider docs and console before every production cost claim.