Picture MeHome

How to Use Nano Banana Pro

Learn how to use Nano Banana Pro (Gemini 3 Pro Image), Google's best AI for images with perfect text, character consistency & advanced editing. Step-by-step tutorial (Nov 2025).

Nano Banana Pro (officially named Gemini 3 Pro Image) is Google's latest AI model for generating and editing images, released on November 20, 2025. It is based on the Gemini 3 Pro model and represents a significant improvement over the original Nano Banana (based on Gemini 2.5 Flash Image). Its key advantages include: superior image text rendering capabilities (supporting clear text in multiple fonts, styles, and languages); more advanced reasoning capabilities for faster understanding; precise editing functions; the ability to simultaneously process and blend multiple people or scenes while maintaining consistency between people and objects; support for resolutions up to 4K; and integration with Google Search to provide factual information (e.g., for generating accurate charts or infographics).

It is not a standalone app or tool but is integrated into Google's ecosystem. As an AI expert, I'll provide an objective, step-by-step tutorial based on official Google documentation and announcements as of November 20, 2025.

1. What Nano Banana Pro Actually Is

  • Full technical name: Gemini 3 Pro Image
  • Codename used inside the Gemini app: Nano Banana Pro (with the 🍌 icon)
  • Built on Gemini 3 Pro (the same reasoning backbone as Gemini 3 Pro text)
  • Designed for higher-quality image generation, advanced editing, perfect text rendering, and stronger real-world knowledge than the previous Nano Banana (Gemini 2.5 Flash Image).

1. Where to Access Nano Banana Pro

The primary (and easiest) way for most users is through the Gemini app or gemini.google.com (Switch the model from “Fast” to “Thinking” → you are now using Nano Banana Pro):

  • Free users: Limited daily generations with Nano Banana Pro (after limit, switches to original faster Nano Banana).
  • Google AI Pro / Ultra subscribers (paid plans): Much higher limits, no visible watermark on images.

Gemini Homepage

Other places (rolling out):

  • Google AI Studio (model name: gemini-3-pro-image-preview)
  • Google Slides and Google Vids (Workspace customers)
  • Google Ads (global)
  • NotebookLM (subscribers)
  • Flow (Google AI Ultra subscribers only)
  • Gemini API and Vertex AI (developers and enterprises)

Free users: small daily quota of Nano Banana Pro generations → then automatically switches back to the faster original Nano Banana. Google AI Pro subscribers: higher quota, images still carry the visible Gemini sparkle watermark. Google AI Ultra and Google AI Studio: no visible watermark (only invisible SynthID).

2. How to Activate Nano Banana Pro in the Gemini App / Website

  1. Open the Gemini app (iOS/Android) or go to https://gemini.google.com.
  2. Start a new chat or continue an existing one.
  3. Tap/click the 🍌 Create images tool (it may appear in the input bar or tools menu).
  4. In the model selector (usually on the right side of the prompt box), choose “Thinking” instead of “Fast”.
    • “Thinking” = Nano Banana Pro (Gemini 3 Pro Image) – slower but higher quality, better reasoning.
    • “Fast” = original Nano Banana (Gemini 2.5 Flash Image) – quicker, lower quality.
  5. You’re now using Nano Banana Pro!

3. Basic Image Generation (Text-to-Image)

  1. Type a descriptive prompt, e.g.:
    • “A photorealistic mountain landscape at sunset with the text ‘Explore More’ in elegant serif font overlaid”
    • “An infographic explaining how solar panels work, clean minimalist style, accurate labels”
  2. Optionally specify aspect ratio or resolution: add phrases like “16:9 aspect ratio” or “4K resolution”.
  3. Hit send → Gemini will think for a few seconds (reasoning step) then generate 2–4 variants.
  4. You can immediately react with thumbs up/down or ask for variations.

Tip: Nano Banana Pro excels at text inside images. Be specific about font, placement, language, etc.

You can then continue the conversation with plain-language edits:

  • “Turn this into nighttime”
  • “Focus on the flowers and add bokeh”
  • “Translate all text to Spanish”
  • “Combine these 12 photos into one group portrait, keep everyone identical”

4. Image Editing & Remixing

Nano Banana Pro’s biggest strength is iterative, context-aware editing.

Method A: Upload an image + text prompt

  1. Click the upload/image icon in the prompt bar.
  2. Upload one or more photos (up to 14 reference images supported).
  3. Describe changes in natural language:
    • “Change the background to a snowy mountain, keep the people exactly the same”
    • “Add a red sports car in the foreground, photorealistic, evening lighting”
    • “Translate all text to Korean and make the poster in cyberpunk style”
    • “Adjust lighting to golden hour, add depth of field bokeh, shift camera angle to low-angle”
  4. Send → It edits while preserving consistency.

Method B: Generate first, then edit in follow-up messages

  • After an image is generated, simply reply:
    • “Make the sky stormy and add rain”
    • “Remove the hat from the person on the left”
    • “Change the color grading to vintage film look”

Pro tip: Use precise language for localized edits: “Only change the shirt to blue, keep everything else identical”.

5. Advanced Features & Prompting Tips

  • Blend multiple images: Upload several photos and prompt “Combine these 5 character references into one consistent group photo on a beach”.
  • Character consistency: “Keep the same 3 people from the reference photos, place them in a futuristic city”.
  • Factual / Search-grounded images: Pro/Ultra users get real-time Search integration:
    • “Create an accurate diagram of the water cycle with labels”
    • “Infographic of today’s weather in Tokyo with icons”
  • Professional controls:
    • Camera: “low angle shot”, “wide-angle lens”, “macro close-up”
    • Lighting: “dramatic rim lighting”, “soft morning light”
    • Style: “in the style of Studio Ghibli”, “photorealistic product render”
    • Text: “Add the slogan ‘Hello World’ in bold sans-serif font at the bottom”
  • Resolution & aspect ratio: Add “4K”, “square”, “vertical 9:16 for stories”, etc.

6. Known Limitations (Google’s Own Words)

  • Can still misspell words, struggle with very small faces or extremely fine details
  • Complex blends or extreme lighting changes may occasionally show artifacts
  • Factual infographics and data visualisations should always be double-checked
  • Translation is very good but not perfect for every idiom or cultural nuance

7. For Developers (Google AI Studio / API)

  1. Go to https://aistudio.google.com
  2. Select model → gemini-3-pro-image-preview (this is Nano Banana Pro)
  3. Use the same text + image prompt interface or API calls for programmatic access.

That’s it! Nano Banana Pro is currently one of the most capable publicly available image models, especially for text-heavy or consistency-required work. Start simple, iterate with follow-up prompts, and you’ll get impressive results quickly.