Legible text and in-image translation
Generates accurate, readable text for mockups, posters and cards, and can translate or localize the text inside an existing image.
Photoreal portraitGenerated image will be displayed here
Google · Gemini 3.1 Flash Image
Nano Banana 2 is Google's Gemini 3.1 Flash Image model, released in February 2026 to bring Nano Banana Pro-level world knowledge and quality to Flash speed. On Motixo you can prompt it directly, add reference images for consistent products and characters, and export 1K, 2K or 4K images in 14 aspect ratios.
Specs from Google's Nano Banana 2 documentation and Motixo's live model settings.
Google describes Nano Banana 2 as Pro-level image generation and editing at Flash speed. These are the capabilities it highlights, and how they show up when you work in Motixo.
Generates accurate, readable text for mockups, posters and cards, and can translate or localize the text inside an existing image.
Google reports it keeps up to five characters recognizable and up to 14 objects faithful in one workflow, so series and storyboards stay on model.
Draws on Gemini's world knowledge to depict specific subjects more accurately and to turn notes into diagrams and infographics.
Sticks more closely to long, multi-part prompts, so layout, wording and style requests land the way you wrote them.
On Motixo: 1K, 2K or 4K output, auto or 14 fixed aspect ratios, and JPG or PNG downloads for social, ecommerce and print layouts.
Google cites more vibrant lighting, richer textures and sharper detail than the original Nano Banana, while keeping Flash-tier speed.
Nano Banana (Gemini 2.5 Flash Image) launched in August 2025 and Nano Banana Pro (Gemini 3 Pro Image) in November 2025. Here is what changed with Nano Banana 2, based on Google's documentation.
The original Nano Banana outputs 1K images. Nano Banana 2 adds 2K and 4K (plus a 512px tier in Google's API); Motixo offers 1K, 2K and 4K.
Google recommends up to 3 input images for the original model. Nano Banana 2 accepts up to 14 references to carry products, characters and style into one image.
Adds 1:4, 4:1, 1:8 and 8:1 on top of the standard ratios, so tall scroll graphics and wide banners no longer need cropping.
Google positions Pro for high-fidelity work that needs maximum factual accuracy, and Nano Banana 2 for fast generation, precise instruction following and high-volume iteration.
Each scenario maps to a capability Google documents for the model. The photoreal samples below match each workflow brief.
Keep packaging, logos and colors intact while you move a product through new settings.

Write the exact headline into the prompt and get a poster where the words are readable.

Build a sequence of frames where the cast keeps the same faces, outfits and props.

Generate at the ratio each channel needs instead of cropping one square image.

Three steps from prompt to a downloadable image.
Step 1
Write the subject, setting, style, lighting and any exact text in quotes. Upload reference images when a product or character must stay the same.
Step 2
Leave the ratio on auto or pick one of 14 fixed ratios. Use 1K for quick drafts and 2K or 4K for final assets, then choose JPG or PNG.
Step 3
Review the result, change one detail at a time in your prompt and regenerate until it matches the brief, then download.
Nano Banana 2 is Google's Gemini 3.1 Flash Image model, announced on February 26, 2026. It combines the world knowledge and quality of Nano Banana Pro with the speed of Gemini Flash, and is used across the Gemini app, Search, Flow and Google Ads. On Motixo you can use it from the generator on this page.
Go back to the generator, write one prompt or upload a reference, and create your first Nano Banana 2 image.