Nano Banana 2 API
by Google
Pro-level quality at Flash speed.
Generate and edit images with Google's Nano Banana 2 through one unified API. A single REST endpoint, async jobs, and webhooks, with no Google Cloud setup to manage.
- 4K
- Max resolution
- ~12s
- Avg. generation
- 5
- Aspect ratios
- 8 credits
- From
Image · example Playground
Try Nano Banana 2 right here.
Capabilities
What's new in Nano Banana 2.
Pro quality at Flash speed
Built on Gemini 3.1 Flash, it brings Nano Banana Pro's reasoning, world knowledge, and visual fidelity into a faster, cheaper package, and it replaced the original Nano Banana.
Faster and cheaper than Pro
It generates in roughly 4 to 6 seconds at about half the per-image cost of Nano Banana Pro, which suits high-volume and iterative work.
Up to 4K output
It supports resolutions from 512px up to 4096px across 14 aspect ratios, covering everything from social posts to print.
Search grounding and world knowledge
It pulls from Gemini's knowledge base and real-time web search to render specific subjects accurately, and it excels at infographics, notes-to-diagrams, and data visualizations.
Strong multi-subject consistency
It keeps multiple characters and objects consistent across a series of images, alongside accurate multilingual text rendering and translation.
Flexible reasoning and provenance
It offers minimal or high reasoning levels, and every output carries an invisible SynthID watermark plus C2PA Content Credentials.
Showcase
See what Nano Banana 2 can create.
Real outputs generated with Nano Banana 2 on Apiframe, each with the prompt behind it.
A clean data visualization titled Global tea consumption by country, horizontal bars, labeled country names and kilograms per capita, a small legend, Orbit Tea rust-and-navy palette on cream, editorial infographic, all numbers and labels perfectly legible, no decorative clutter.
A 4K product photo of the Orbit Tea ceramic teapot on wet dark slate, round rust-glaze pot with a navy lid, soft morning sidelight, sharp water droplets on the glaze, shallow depth of field, no text, no logo, commercial still life.
Turn handwritten cafe meeting notes into a clean flowchart of the Orbit Tea brewing process: Leaf, Rinse, Steep, Pour, Serve. Five boxes with short labels, rust arrows on cream paper, small Orbit Tea wordmark, all type perfectly legible, diagram style, no misspellings.
A greeting card reading THANK YOU in elegant rust script, watercolor tea-leaf and steam border, cream paper, Orbit Tea small in sans type at the bottom, print-ready, all lettering perfectly legible, no misspellings.
A bilingual cafe menu board for Orbit Tea, English and Japanese on each line: Sencha 煎茶, Hojicha ほうじ茶, Gyokuro 玉露, with yen prices. Hand-lettered chalk on dark slate, modern layout, all text perfectly legible, no misspellings.
An accurate illustrated map infographic of Japan's major tea regions: Shizuoka, Uji, Kagoshima, Sayama. Labeled prefectures, a small key of tea types, Orbit Tea editorial palette rust and navy on cream, all place names correctly spelled, no decorative clutter.
Five named Orbit Tea staff around a round cedar table in a sunlit tea house: Mira (short black bob, rust apron), Kenji (round glasses, navy cardigan), Asha (gold hoop earrings, cream turtleneck), Leo (freckles, rolled chambray sleeves), Noor (silver crop, olive work shirt). Each holds a different ceramic cup from the same glaze family. Warm morning light, plants, a chalkboard menu softly out of focus. Photoreal, consistent faces, no text overlay.
Pricing
Simple per-generation pricing.
8–19 credits per generation ≈ $0.08–$0.19
The rate depends on the resolution and options you pick.
- You only pay for successful generations. Failed jobs are refunded automatically.
- One credit balance across every model on Apiframe, no per-model plans.
- Start with free credits. No subscription required.
- Volume discounts on larger plans.
Specs
At a glance.
Comparison
Nano Banana 2 vs Nano Banana Pro vs GPT Image 2
Nano Banana 2 against the tiers above and beside it. Swap one parameter to move.
| Feature | Nano Banana 2 | Nano Banana Pro | GPT Image 2 |
|---|---|---|---|
| Max resolution | 4K | 4K | High |
| Avg. generation | ~12s | ~20s | ~30s |
| Signature strength | Speed/quality sweet spot | Deepest reasoning + 4K finals | Batch of 8 + text accuracy |
| Price per image | 8–19 credits (≈$0.08–$0.19) | 19–38 credits (≈$0.19–$0.38) | 2–17 credits (≈$0.02–$0.17) |
| Best for | Everyday generation and editing | Complex briefs and hero assets | Instruction-heavy batch work |
Specs reflect what each model supports on Apiframe today, from the live catalog.
Quickstart
How to call the Nano Banana 2 API.
Send one POST /v2/images/generate request with your API key. The call returns a jobId you can poll, or pass a webhook_url and the result is pushed to you the moment it's ready.
Swap nano-banana-2 for any other model and nothing else changes.
const response = await fetch("https://api.apiframe.ai/v2/images/generate", {
method: "POST",
headers: {
"X-API-Key": "afk_your_api_key_here",
"Content-Type": "application/json",
},
body: JSON.stringify({
"prompt": "a sleek silver sports car on a coastal highway at sunset, hyper-realistic",
"model": "nano-banana-2",
"nanoBananaParams": {
"resolution": "1K",
"image_input": "https://example.com/input.jpg",
"output_format": "jpg",
"google_search": false
}
}),
});
const { jobId } = await response.json();
console.log(jobId); Input schema
Every field Nano Banana 2 accepts, with types and defaults. The full reference lives in the docs.
| Field | Type | Description |
|---|---|---|
| prompt required | string | Text description of what to generate. |
| model required | string | The model identifier for this endpoint. Default:"nano-banana-2" · "nano-banana-2" |
| nanoBananaParams.resolution | string | Resolution Default:"1K" · "512px", "1K", "2K", "4K" |
| nanoBananaParams.image_input | string (URL) | Compose with or edit a reference image. |
| nanoBananaParams.output_format | string | Output format Default:"jpg" · "jpg", "png", "webp" |
| nanoBananaParams.google_search | boolean | Google search grounding Default:false |
| nanoBananaParams.image_search | boolean | Image search grounding Default:false |
Nano Banana versions
Every Nano Banana image model on Apiframe. Switch with one parameter.
Use cases
What teams build with Nano Banana 2.
The default for new builds
Near-Pro output at half the price and roughly twice the speed, the tier most teams standardize on for generation and editing.
Generation + editing in one
One model covers text-to-image and conversational editing, so a single integration handles the full image workflow.
World knowledge on tap
Search grounding and multimodal knowledge keep real subjects, products, landmarks, people, accurate instead of approximate.
Multi-subject consistency
Strong identity holding across subjects and edits, up to 4K when the deliverable demands it.
Reviews
9,800+ developers. One API.
Real Apiframe reviews from Trustpilot and G2.
Speeds up our ad creative iteration
The API integrates directly into my existing pipeline, so I can generate and iterate on image variations for Meta ads without jumping between tools. We can go from concept to a testable ad in the same day instead of spreading it across a week.
Fast, reliable, easy to integrate
Image generation requests are processed quickly, and the results are consistently high quality. Customer support has also been responsive whenever I had questions. It has saved me a lot of development time.
Seamless integrations, effortless setup
Generating ad creatives for Facebook and Google ads is more streamlined now since we use APIFRAME to manage the workflow, which used to require a lot of time with official APIs. The initial setup was very easy, just a matter of using curl.
Nano Banana 2 questions
What teams ask before shipping with Nano Banana 2. Everything else lives in the docs.
What is Nano Banana 2?
Google's Gemini 3.1 Flash Image, released in February 2026 to deliver Pro-level quality at Flash speed, replacing the original Nano Banana.
How is it different from Nano Banana Pro?
It is around four times faster and roughly half the cost with comparable quality, while Pro remains the pick for the most demanding text and branding work.
What resolution does it produce?
Up to 4K, with options from 512px to 4096px and 14 aspect ratios.
How fast and how affordable is it?
Around 4 to 6 seconds per image, and on Apiframe from 8 credits ($0.08) at 1K to 19 credits ($0.19) at 4K, about half the price of Pro.
How many subjects can it keep consistent?
Multiple characters and objects across images, supporting consistency for several characters and many objects at once.
Where can you access it?
Through Apiframe, as well as the Gemini app, Google AI Studio, the Gemini API, and Vertex AI.
Related models
All image modelsFlux 2 Max
by Black Forest Labs
Highest-quality Flux 2 tier.
View modelMidjourney
by Midjourney
Iconic stylised, cinematic look.
View modelSeedream 5.0 Pro
by ByteDance
Flagship ByteDance model with sharp 1K/2K output and up to 10 reference images.
View modelIdeogram v4 Quality
by Ideogram
Highest-quality Ideogram v4 tier for production-ready images.
View model9,800+
developers & companies
Official partner of
Power your next AI product with Apiframe.
Instant access to 70+ media models through a single API. Start free and scale when you're ready.