Google released Nano Banana 2.1 on October 6, 2026, built on Gemini 3.6 Flash to bring text and reference images together in one model. New accounts get 20 free credits on sign-up, no card required.
GALLERY
Feature 01
Nano Banana 2.1 keeps distinct subjects recognizable when you place multiple people or assets into a new environment. In Google's tests, it scored 1106 in multi-character consistency compared to 978 for Nano Banana 2 and 1011 for Nano Banana Pro. You can upload reference pictures to hold character traits steady across images. While character consistency between the input image and output is not always perfect, it handles multi-character setups better than previous generations.
Feature 02
Combine style, subject, and background elements from different source photos into a single coherent image. In Google's tests, Nano Banana 2.1 scored 1066 for multi-reference editing compared to 988 for Nano Banana 2 and 989 for Nano Banana Pro. You can upload up to 8 reference images on RiceBowl to guide the model's subject placement and overall scene layout. It gives you a practical way to merge separate photos without starting your prompt from scratch.
Feature 03
Nano Banana 2.1 lays out infographics more cleanly and gets the facts in them right more often than older models. In Google's tests, the model reached an infographic factuality accuracy score of 0.521 compared to 0.179 for Nano Banana 2 and 0.265 for Nano Banana Pro. It also scored 1048 for infographic design, higher than Nano Banana 2 at 961 and Nano Banana Pro at 912. Google notes that small text can come out blurry, especially at 1K resolution, and long paragraphs remain a weak point.
Feature 04
Transform real photos into distinct artistic styles like watercolors, sketches, or oil paintings while maintaining basic composition. In Google's tests, it achieved a 1062 stylization score, outperforming Nano Banana 2 at 991 and Nano Banana Pro at 990. You can apply stylistic changes using plain text prompts or style reference images in the RiceBowl generator. The model adapts lighting, textures, and brushwork to match your target visual aesthetic.
PROMPT GUIDE
Prompt
Put the fox and the rabbit from the reference image in a cozy bookstore on a rainy evening: the rabbit reads a book aloud from a rolling ladder, the fox listens from an armchair holding a mug. Keep both characters' faces, colors and outfits the same as the reference. 3D animated style, warm lamp light, rain on the window behind them.
Input
Output
Prompt
Clean flat infographic about houseplant watering on a white background. Title at the top in bold green sans-serif: "Water Less, Grow More". Three columns below, each with a simple plant icon and a short label: "Snake Plant: every 3 weeks", "Pothos: every week", "Fern: every 3 days". A thin divider line and a footer note in grey: "Check the top inch of soil first".
Output
Prompt
Turn this photo into a claymation scene with soft modeling-clay textures and visible fingerprints. Keep the dog's pose, the blanket and the park background in the same places.
Input
Output
USE CASES
Place products into clean studio setups or lifestyle backgrounds using text prompts and reference photos. In Google's tests, the model scored 1024 for product consistency compared to 955 for Nano Banana 2. You can upload product shots to preserve item details across different lighting conditions and staging environments.
Generate banners, promotional posts, and eye-catching artwork tailored to specific aspect ratios. The model handles varied artistic styles and prompt instructions directly in your browser. Pick an aspect ratio such as 1:1 or 9:16 on RiceBowl to fit feed posts or stories.
Maintain character traits across sequential frames to plan comics or video scenes. The model scored 1106 in multi-character consistency in Google's tests, beating older models when placing recurring figures into new settings. Reference uploads help keep character features recognisable from frame to frame.
Combine concise text, clear diagrams, and custom graphics for educational or marketing materials. The model scored 1048 for infographic design and 0.521 for factuality accuracy in Google's tests. It helps build structured poster layouts directly from text prompts and design references.
BALANCED REVIEW
MODEL FEATURES
01
The model is built on Google's Gemini 3.6 Flash architecture, part of the Gemini 3 family. It processes text and images together through a context window of up to 1M tokens to produce generated images and accompanying text outputs.
02
Google evaluates two modes, Thinking and No Thinking; Thinking scores higher on overall preference, character consistency, and infographic design.
03
RiceBowl lets you run Nano Banana 2.1 in your browser without installing software or managing API keys. Upload up to 8 reference images, pick 1K, 2K, or 4K resolution, and generate new visuals in a single workspace.
BUYER'S GUIDE
Choose Nano Banana 2.1. In Google's tests, it outperforms Nano Banana Pro across overall preference (1050 vs 935), multi-character consistency (1106 vs 1011), and infographic accuracy (0.521 vs 0.265). It represents a clear upgrade over Pro for editing, stylization, and layout tasks.
Yes, the new model offers measurable improvements over Nano Banana 2. Google's test scores show gains in multi-reference editing (1066 vs 988), infographic design (1048 vs 961), and general editing (1026 vs 938). If you build complex graphics or edit multi-source visuals, the newer model delivers better results.
Avoid it if your project depends heavily on rendering long paragraphs of text or flawless spatial precision. Google notes that long text blocks remain weak and the model can mix up left and right orientations. For dense typographic documents or exact positional edits, dedicated graphic design software is a safer choice.
BENCHMARKS
| Google test | Nano Banana 2.1 (Thinking) | Nano Banana 2 | Nano Banana Pro |
|---|---|---|---|
| Overall preference | 1050 | 990 | 935 |
| Multi-character consistency | 1106 | 978 | 1011 |
| Multi-reference editing | 1066 | 988 | 989 |
| Stylization | 1062 | 991 | 990 |
| Infographic design | 1048 | 961 | 912 |
| Infographic factuality | 0.521 | 0.179 | 0.265 |
| General editing | 1026 | 938 | 939 |
| Product consistency | 1024 | 955 | 965 |
Source: Google DeepMind Nano Banana 2.1 model card (Oct 6, 2026). Nano Banana 2.1 scores are Thinking mode; all rows are Elo except infographic factuality (accuracy).
Try It Free01
Type a detailed description of the scene, subject, or image modifications you want to generate in the text box. Select your desired aspect ratio and choose output resolutions of 1K, 2K, or 4K.
02
Optionally upload up to 8 reference images to guide character features, styles, or overall composition. The model processes your text and visual references together in a single prompt.
03
Click Generate to run the model directly inside your web browser. Your generated output will appear in the feed below, ready for instant preview and download.
One balance, every model. Credits work across all 20 video models — no watermark, no tier-locked models, commercial use on every paid plan.
Save 30% billed annually
Flexible monthly subscription.
The lowest price to get started. Every model, a monthly budget to try things out.
$7.99/mo
Renews monthly, cancel anytime.
200 credits
≈ 0–10 videos, depending on the model
≈ 14–66 images, depending on the model
4.00¢ / credit
The cheapest way to start. Every model, a small monthly budget.
$12.99/mo
Renews monthly, cancel anytime.
400 credits
≈ 1–20 videos, depending on the model
≈ 28–133 images, depending on the model
3.25¢ / credit
Enough credits to try every model and finish a few clips each month.
$29.9/mo
Renews monthly, cancel anytime.
1,000 credits
≈ 3–50 videos, depending on the model
≈ 71–333 images, depending on the model
2.99¢ / credit
2 parallel tasks
Steady output for solo creators going pro — every model, a comfortable budget.
$49.9/mo
Renews monthly, cancel anytime.
2,000 credits
≈ 6–100 videos, depending on the model
≈ 142–666 images, depending on the model
2.50¢ / credit
3 parallel tasks
High-volume production for prolific creators and small teams.
$89.9/mo
Renews monthly, cancel anytime.
5,000 credits
≈ 15–250 videos, depending on the model
≈ 357–1666 images, depending on the model
1.80¢ / credit
5 parallel tasks
Pay yearly and save more.
Enough credits to try every model and finish a few clips each month.
$19.9/mo$29.9
Billed yearly at $238.8. Renews yearly, cancel anytime.
Save $120/yr vs monthly
12,000 credits
≈ 38–600 videos, depending on the model
≈ 857–4000 images, depending on the model
1.99¢ / credit
2 parallel tasks
Steady output for solo creators going pro — every model, a comfortable budget.
$34.9/mo$49.9
Billed yearly at $418.8. Renews yearly, cancel anytime.
Save $180/yr vs monthly
24,000 credits
≈ 76–1200 videos, depending on the model
≈ 1714–8000 images, depending on the model
1.75¢ / credit
3 parallel tasks
High-volume production for prolific creators and small teams.
$62.9/mo$89.9
Billed yearly at $754.8. Renews yearly, cancel anytime.
Save $324/yr vs monthly
60,000 credits
≈ 190–3000 videos, depending on the model
≈ 4285–20000 images, depending on the model
1.26¢ / credit
5 parallel tasks
Top up credits valid for 12 months.
One-time top-up — perfect for occasional needs.
$49.9
One-time purchase, valid 12 months.
1,000 credits
≈ 3–50 videos, depending on the model
≈ 71–333 images, depending on the model
4.99¢ / credit
Larger top-up with 20% bonus — best per-credit value for one-time buyers.
$198
One-time purchase, valid 12 months.
4,800 credits
≈ 15–240 videos, depending on the model
≈ 342–1600 images, depending on the model
4.13¢ / credit
5 parallel tasks
Nano Banana 2.1 is Google's new image generation and editing model released on October 6, 2026. Built on Gemini 3.6 Flash, it processes text and images together through a context window of up to 1M tokens to output images and text.
Yes, you can try it free on RiceBowl AI. New accounts get 20 free credits on sign-up, no card required. Each image you generate uses credits from your balance.
The new model outperforms Nano Banana Pro across Google's preference tests, scoring 1050 in overall preference compared to Pro's 935. It also shows marked improvements in multi-character consistency (1106 vs 1011) and infographic factuality (0.521 vs 0.265).
Yes, in Google's tests the new model achieved higher Elo scores than Nano Banana 2 in overall preference (1050 vs 990), multi-reference editing (1066 vs 988), and general image editing (1026 vs 938). It also leads in multi-character consistency (1106 vs 978) and stylization (1062 vs 991).
Google offers Nano Banana 2.1 in the Gemini app, Google AI Studio, the Gemini API, AI Mode in Google Search, Google Ads, Google Flow, and Google Stitch. On RiceBowl, you can run it instantly in your browser without any setup.
Yes, Google offers access to the model through the Gemini API and Google AI Studio for developer integrations. On RiceBowl, you can generate images with it directly in your browser without needing an API key.
On RiceBowl, you can upload up to 8 reference images per generation to guide the model. This allows you to combine styles, characters, and backgrounds into a single visual output.
Google lists several known limitations for the model, including blurry small text at 1K resolution and weak rendering of long paragraphs. Additionally, character consistency from input images is not always perfect, doodle or mask edits only partly follow instructions, spatial left/right mix-ups can occur, and world knowledge remains limited.
Yes. Images you generate on RiceBowl can be used commercially for your projects or business materials. If you generate images in Google's own apps instead, Google's terms apply there.
RELATED
MORE MODELS
Experience Google's newest image model in your browser with no installation needed.
Generate with Nano Banana 2.1 FreeStart Generating with Nano Banana 2.1 Today
Try It Free