Nano Banana 2.1 is an update to Nano Banana 2 (Gemini 3.1 Flash Image) for fast image generation and conversational editing, with improved visual quality, text rendering, multi-turn consistency and Google Search grounding at 1K, 2K and 4K.
| Developer | |
|---|---|
| Model ID | google/gemini-nano-banana-2.1 |
| Modalities | Text → Image |
| Resolutions | 1K, 2K, 4K |
| Released | 6 October 2026 |
| Endpoint | POST /v1/images/generations |
Conversational editing
An image can be adjusted across turns rather than regenerated from scratch each time.
Text rendering
Improved over Nano Banana 2, which is the usual failure point in generated design work.
Search grounding
Google Search grounding for subjects that have to match something real.
Resolution you choose
1K, 2K and 4K output, billed at the standard rate, 1.5x and 2x respectively.
Design with text in it
Posters, covers and ads where legible typography decides whether the asset is usable.
Iterative work
Multi-turn consistency keeps a look stable while a detail is refined over several passes.
Sources: https://ai.google.dev/gemini-api/docs/models · https://fal.ai/models/google/nano-banana-2.1
Measured on the AI/ML API gateway across live traffic. Figures are medians unless stated. Last updated 2026-10-12.
import requests r = requests.post( "https://api.aimlapi.com/v1/images/generations", headers={"Authorization": "Bearer " + AIMLAPI_KEY}, json={ "model": "google/gemini-nano-banana-2.1", "prompt": "A weathered brass compass resting on a nautical chart", "resolution": "2K" }, ) print(r.json())
const r = await fetch("https://api.aimlapi.com/v1/images/generations", { method: "POST", headers: { Authorization: "Bearer " + process.env.AIMLAPI_KEY, "Content-Type": "application/json", }, body: JSON.stringify({ "model": "google/gemini-nano-banana-2.1", "prompt": "A weathered brass compass resting on a nautical chart", "resolution": "2K" }), }); console.log(await r.json());
curl -X POST https://api.aimlapi.com/v1/images/generations \ -H "Authorization: Bearer $AIMLAPI_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"google/gemini-nano-banana-2.1","prompt":"A weathered brass compass resting on a nautical chart","resolution":"2K"}'
OpenAI-compatible — swap the base URL and it works with your existing SDK.
| Type | Price |
|---|---|
| Input | |
| Output | |
Billed per generation. 2K output is charged at 1.5x the standard rate, 4K at 2x, and 0.5K at 0.75x.
| Benchmark | Score | What it measures | Source | Retrieved |
|---|---|---|---|---|
| Text-to-Image Arena | 1160 | Human preference Elo from blind pairwise comparisons of generated images, measured independently by Artificial Analysis | Source | October 7, 2026 |
| Model | Input | Output | Context | Best for |
|---|---|---|---|---|
Nano Banana 2.1 This page | Image generation | |||
| Precise image editing | ||||
| Image generation at scale | ||||
| Image generation at scale |
Nano Banana 2.1 is Google's image model. It generates images from text prompts, and Google describes it as a high-efficiency image generation and editing model with improved visual quality, typography and multi-turn consistency.
0.5K (512px), 2K and 4K, set through the resolution parameter. Aspect ratio is adjustable.
$0.104 per generation at the standard rate. 2K output is charged at 1.5x that rate, 4K at 2x, and 0.5K at 0.75x.
Yes, web search grounding is an optional parameter, as is high thinking. Both are off by default.
Nano Banana 2.1 generates an image from a text prompt alone. Nano Banana 2.1 Edit takes reference images and transforms them from text instructions. They are separate endpoints at the same price.
