Introducing Google Nano Banana 2 Edit on WaveSpeedAI

Google’s Nano Banana 2 has arrived — and with it, one of the most capable AI image editing models ever released. Nano Banana 2 Edit (powered by Gemini 3.1 Flash Image) brings natural language image editing to a new level: describe the change you want, and the model executes it with precision, context awareness, and up to 4K output resolution.

Whether you need to swap a background, translate text within an image, change a character’s appearance, or apply a completely new visual style, Nano Banana 2 Edit handles it through simple text instructions — no masks, no layers, no manual selection tools.

Try Nano Banana 2 Edit on WaveSpeedAI

What Is Nano Banana 2 Edit?

Nano Banana 2 Edit is the image-to-image editing variant of Google’s latest Nano Banana model family. Built on the Gemini 3.1 Flash architecture, it combines the visual intelligence and quality of Nano Banana Pro with the speed of Google’s Flash tier — delivering what Google calls “Pro quality at Flash speed.”

Unlike traditional image editors that require pixel-level manipulation, Nano Banana 2 Edit understands the semantic content of your images. It knows what’s a person, what’s a background, what’s text, and what’s a shadow. When you ask it to “change the background to a sunset beach,” it doesn’t just crop and paste — it adjusts lighting direction, color temperature, and reflections to make the edit look natural.

In real-world production environments, this approach delivers dramatic efficiency gains. HubX, an early adopter, reported a 74-76% reduction in latency after integrating Gemini 3.1 Flash Image into their face editing workflows — making edits roughly 4x faster without sacrificing Pro-level quality.

Key Features

Natural Language Editing

Describe your edits in plain English. The model interprets context, spatial relationships, and visual semantics to execute complex transformations:

“Change the man’s suit from navy to charcoal gray”
“Replace the coffee cup with a glass of orange juice”
“Make the scene look like golden hour lighting”
“Add rain to the window in the background”

The more specific your instructions, the more precise the results. But even broad instructions like “make this look more professional” produce meaningful improvements.

Multi-Image Reference

Upload up to 14 reference images for complex edits and compositions. This enables workflows like:

Style transfer from a reference photo
Character consistency across multiple edited outputs
Compositing elements from several source images into one coherent scene

Nano Banana 2 Edit maintains the fidelity of referenced objects and can preserve resemblance across up to 5 characters in a single editing workflow.

4K Output Resolution

Generate edited images at 1K, 2K, or 4K resolution — covering everything from quick preview iterations to production-ready assets. Start with 1K for rapid experimentation, then regenerate your best results at 4K for final delivery.

Flexible Aspect Ratios

Output in any of 10 supported aspect ratios: 1:1, 3:2, 2:3, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, and 21:9. This covers the full range of modern content formats — from square social posts to ultrawide cinematic banners.

In-Image Text Editing and Translation

One of Nano Banana 2 Edit’s standout capabilities is its handling of text within images. The model can:

Modify existing text while preserving the surrounding design
Translate text into different languages within the same image
Maintain font style, color, and layout consistency after edits

This is a historically weak point for AI image models, and Nano Banana 2 represents a generational leap in text rendering accuracy.

Built-In Prompt Enhancer

Not sure how to phrase your edit? The built-in prompt enhancer automatically refines your descriptions for better results — turning vague instructions into precise, model-optimized prompts.

Real-World Use Cases

E-Commerce Product Photography

Transform basic product shots into polished catalog images. Change backgrounds, adjust lighting, swap colors, or place products in lifestyle settings — all without reshooting. A single product photo can be edited into dozens of variants for A/B testing or regional campaigns.

Marketing and Advertising

Quickly iterate on visual concepts. Change headline text, swap model appearances, adjust brand colors, or localize campaign imagery for different markets. The text translation capability is particularly valuable for global campaigns — edit the text in your hero image from English to Japanese, Spanish, or any target language while preserving the design.

Content Creation at Scale

Social media managers and content teams can transform a single base image into platform-specific variations. Adjust aspect ratios, modify visual elements for different audiences, and maintain brand consistency across dozens of edited outputs.

Character and Concept Design

Game studios and creative agencies can use multi-image reference editing to iterate on character designs — changing outfits, accessories, expressions, or environments while maintaining character identity across variations.

Photo Restoration and Enhancement

Upgrade casual photos to professional quality. Fix lighting issues, remove distracting elements, change weather conditions, or adjust the time of day in outdoor shots. The model preserves authentic textures while enhancing overall quality.

Getting Started on WaveSpeedAI

Using Nano Banana 2 Edit through WaveSpeedAI’s API is straightforward:

import wavespeed

output = wavespeed.run(
    "google/nano-banana-2/edit",
    {
        "prompt": "Change the background to a modern office with floor-to-ceiling windows",
        "image": "https://example.com/portrait.jpg",
    },
)

print(output["outputs"][0])

For higher resolution output, add the resolution parameter:

import wavespeed

output = wavespeed.run(
    "google/nano-banana-2/edit",
    {
        "prompt": "Replace the text on the sign with 'GRAND OPENING' in bold red letters",
        "image": "https://example.com/storefront.jpg",
        "resolution": "4k",
    },
)

print(output["outputs"][0])

Pricing

Nano Banana 2 Edit delivers Pro-level editing quality at Flash-tier pricing:

Resolution	Cost per Edit
1K	$0.08
2K	$0.12
4K	$0.16

With generation times in the 4-6 second range and no cold starts on WaveSpeedAI, you get fast, predictable performance at a fraction of what manual editing workflows cost.

How It Compares

Nano Banana 2 Edit occupies a unique position in the AI editing landscape. Compared to general-purpose image generators like Midjourney or DALL-E, it offers significantly more precise control over edits — preserving the parts of your image you want to keep while transforming only what you specify. And compared to Nano Banana Pro Edit, it delivers comparable quality at faster speeds and lower cost.

Feature	Nano Banana 2 Edit	Nano Banana Pro Edit
Architecture	Gemini 3.1 Flash	Gemini 3 Pro
Max Resolution	4K	4K
Reference Images	Up to 14	Up to 14
Speed	Fast (4-6s)	Moderate
Text Editing	Advanced	Advanced
Cost (1K)	$0.08	Higher

For most editing workflows, Nano Banana 2 Edit is the optimal choice — matching Pro-level output quality while keeping costs and latency low enough for high-volume production use.

Start Editing

Nano Banana 2 Edit makes professional image editing as simple as describing what you want. No manual masking, no layer management, no design skills required — just clear instructions and the model handles the rest.