Nano Banana Pro is Google's reasoning-first image model, available on Higgsfield. It analyzes your prompt for logic, quantities, and spatial layout before rendering: use it when precision matters, the correct number of objects, readable text, consistent faces, and accurate perspective. For faster, higher-volume production at comparable quality, use Nano Banana 2, and for edits to a specific area, use Nano Banana Pro Inpaint.
What is the Nano Banana family?
Model | Best for | Key detail |
|---|---|---|
Nano Banana | Edit an existing image with precise instructions | Up to 8 reference images |
Nano Banana Apps | One-click style transformations | 2,000+ ready-made style apps |
Nano Banana Pro | Complex scene generation with reasoning | Analyzes the prompt before rendering |
Nano Banana Pro Inpaint | Edit a specific area without changing the rest | Paint-to-edit, structure-preserving |
Nano Banana 2 | Fast bulk generation at Pro-level quality | Up to 5 characters, native 4K |
What makes Nano Banana Pro different?
Nano Banana Pro analyzes the prompt first: it evaluates relationships between objects, checks for logical consistency, and interprets quantities, spatial layout, and lighting before rendering. This solves problems that trip up most image models: the wrong number of objects, broken physics, distorted text, inconsistent faces, and perspective errors.
Use Nano Banana Pro when your prompt has a specific number of objects or people, when you need readable text in the image, when identity consistency matters, or when you're generating product visuals where errors are costly. Use Nano Banana 2 when you need the same quality faster and at higher volume, or when your scene has multiple characters (up to 5) or needs native 4K output.
How do I generate with Nano Banana Pro?
Open higgsfield.ai → Image → Nano Banana Pro.
Choose your input mode: Text-to-Image, Multi-reference (up to 8 reference images), or Image-to-Image.
Write a clear, specific prompt. Describe the subject, setting, lighting, and any specific quantities or text. Be explicit.
Add references if needed, and describe the role of each reference clearly in the prompt.
Set the resolution. Start at 1K when iterating, scale to 2K or 4K for final assets.
Generate and iterate. If something is off, change the prompt or the reference, not both at once.
How do I use Nano Banana Pro Inpaint?
Open higgsfield.ai → Edit → Nano Banana Pro Inpaint.
Upload your image.
Paint the area you want to change.
Write a short instruction describing the change.
Generate.
The model rebuilds only the masked region, matching the original image's lighting, shadows, textures, and depth of field. Best for retouching portraits, editing product labels, fixing faces or artifacts, and replacing objects.
How do references work?
The standard Nano Banana model accepts up to 8 reference images in one composition, and each reference can serve a different role: describe what to take from each in your prompt.
For consistent character identity across multiple separate generations, save the character as an Element: click @ Elements under the prompt field (or type @ in the prompt) and select it, with no re-uploading per image. If the character is a real person, train a Soul ID first and create the Element from its portraits. See How do I create and use a Soul ID character?
Common mistakes and fixes
Output ignores a specific quantity or layout. Be more explicit: "exactly 3 bottles on the left side of the frame" rather than "a few bottles."
Text in the image is blurry or misspelled. Put the exact text in quotes in your prompt, and describe the font style, size, and placement explicitly.
Output looks blurry at 2K or 4K. Retry the generation. If the issue persists, contact us through higgsfield.ai/contact with the generation ID: you can copy it from the generation's page URL in Assets.
Reference image is rejected. Check that the file is JPG, PNG, or WebP and within the size limit, and avoid images with heavy compression or watermarks.