Gemini Nano Banana 2.1: Specs, API Price per Image and How to Use It
Gemini Nano Banana 2.1 (gemini-nano-banana-2.1) is Google's latest fast image generation and editing model, released October 6, 2026 as an update to Nano Banana 2 (Gemini 3.1 Flash Image). It improves image quality, prompt adherence, character consistency and text rendering, outputs 1K, 2K or 4K images, and accepts up to 14 reference images. On the Gemini API it costs $0.0336 per 1K image, $0.0504 per 2K image and $0.113 per 4K image (half that with the Batch API). There is no free API tier for it, but Google lists it as testable in Google AI Studio. It is cheaper per image than Nano Banana 2 at every resolution.
Nano Banana 2.1 at a glance
| Gemini Nano Banana 2.1 | |
|---|---|
| Model code | gemini-nano-banana-2.1 |
| Released | October 6, 2026 (stable) |
| What it replaces | An update to Nano Banana 2 (Gemini 3.1 Flash Image); the faster, cheaper counterpart to Nano Banana Pro (Gemini 3 Pro Image) |
| Input | Text, images, video and PDF (up to 131,072 tokens) |
| Output | Images and text (up to 32,768 tokens) |
| Resolutions | 1K (default), 2K, 4K |
| Reference images | Up to 14 (consistency for up to 4 characters and up to 10 objects) |
| Thinking | Configurable: minimal, medium (default), high |
| Grounding | Google Web and Image Search |
| Batch API | Supported |
| Watermark | Every generated image carries Google's SynthID watermark |
Source: Gemini API model page and image generation docs.
What's new in Nano Banana 2.1
From Google's model page:
- Better visual quality and realism at 1K, 2K and 4K.
- Fixed tiling artifacts on very wide and tall aspect ratios (1:4, 4:1, 1:8, 8:1) at 2K and 4K.
- Sharper text rendering and more accurate infographic layouts.
- Multi-image fusion from up to 14 reference images.
- Search grounding with Google Web and Image Search.
- Thinking levels you can set (minimal / medium / high).
Nano Banana 2.1 API price per image
Google charges by token. One output image uses a fixed number of tokens, so the price per image is fixed:
| Resolution | Tokens per image | Standard price | Batch API price |
|---|---|---|---|
| 1K (1024×1024) | 1,120 | $0.0336 | $0.0168 |
| 2K (2048×2048) | 1,680 | $0.0504 | $0.0252 |
| 4K (4096×4096) | 3,780 | $0.113 | $0.0567 |
Other rates (per 1M tokens, standard): input $1.50 (text/image/video), text and thinking output $7.50, image output $30. Batch rates are half. Search grounding includes 5,000 free search requests per month (shared across Gemini 3 and newer models), then $14 per 1,000. Source: Gemini API pricing, checked October 18, 2026.
What 100 or 1,000 images cost
| Images | 1K | 2K | 4K |
|---|---|---|---|
| 100 (standard) | $3.36 | $5.04 | $11.30 |
| 1,000 (standard) | $33.60 | $50.40 | $113.00 |
| 1,000 (Batch API) | $16.80 | $25.20 | $56.70 |
Input tokens (your prompt and reference images) are billed on top, at $1.50 per 1M.
Nano Banana 2.1 vs Nano Banana 2 price
| Per image (standard) | Nano Banana 2.1 | Nano Banana 2 (Gemini 3.1 Flash Image) |
|---|---|---|
| 1K | $0.0336 | $0.067 |
| 2K | $0.0504 | $0.101 |
| 4K | $0.113 | $0.151 |
| Image output per 1M tokens | $30 | $60 |
| Input per 1M tokens | $1.50 | $0.50 |
Nano Banana 2.1 costs about half as much per 1K or 2K image as Nano Banana 2, but its input tokens cost more. Heavy reference-image workflows should compare total cost, not just output.
Is Nano Banana 2.1 free?
- Gemini API: no. Google lists the free tier as "not available" for this model.
- Google AI Studio: Google marks the model as testable in AI Studio.
- Gemini app: launch coverage reports that Nano Banana 2.1 is the default image model for Google AI Plus, Pro and Ultra subscribers, while free Gemini users stay on an older Nano Banana model. Check the model name shown in the Gemini app.
How to use Nano Banana 2.1 with the API
Get an API key from Google AI Studio, then (Python, from Google's docs):
from google import genai
import base64
client = genai.Client()
interaction = client.interactions.create(
model="gemini-nano-banana-2.1",
input="A nano banana dish in a fancy restaurant, Gemini theme",
)
Set the resolution (1K, 2K or 4K) and aspect ratio in the image response format; the image generation docs show both. For edits, send the follow-up request with the previous interaction id ("Translate this infographic to Spanish, change nothing else").
Prompt tips that match 2.1's strengths
- Text in images: put the exact words in quotes and say where they go ("headline 'Fresh Daily' at the top").
- Characters: attach up to 4 character reference images and name them in the prompt to keep them consistent across shots.
- Products: attach object references (up to 10) for logos and packaging you need to stay exact.
- Panoramas and banners: wide ratios like 4:1 or 8:1 at 2K/4K no longer tile, so use them for headers.
- Facts on infographics: turn on search grounding so labels and data are checked against Google Search.
Frequently asked questions
What is Gemini Nano Banana 2.1?
Google's latest fast image generation and editing model (gemini-nano-banana-2.1), released October 6, 2026 as an update to Nano Banana 2 (Gemini 3.1 Flash Image), with better quality, prompt adherence, character consistency and text rendering.
How much does Nano Banana 2.1 cost per image?
On the Gemini API: $0.0336 per 1K image, $0.0504 per 2K image and $0.113 per 4K image. The Batch API halves those prices. Input tokens cost $1.50 per 1M.
Is Nano Banana 2.1 free?
Not on the Gemini API: Google lists no free tier for this model. Google marks it as testable in Google AI Studio, and reports say it is the default in the Gemini app for paid Google AI plans.
What is the Nano Banana 2.1 model ID?
gemini-nano-banana-2.1.
Is Nano Banana 2.1 better than Nano Banana 2?
Google describes it as an update with better visual quality, prompt adherence, multi-turn character consistency and text rendering, plus fixed tiling on wide aspect ratios. It is also about half the price per 1K or 2K image.
What resolutions does Nano Banana 2.1 support?
1K (default), 2K and 4K.
How many reference images can Nano Banana 2.1 use?
Up to 14, with consistency for up to 4 characters and up to 10 objects.
Does Nano Banana 2.1 add a watermark?
Yes. Google adds an invisible SynthID watermark to every generated image.