← Models

Nano Banana 2

Google · Nano Banana

Google's latest fast image-generation release combining higher fidelity, stronger reasoning, and Flash-speed iteration.

Type
image
Context
N/A
Max Output
N/A
Status
preview
API Access
Yes
License
proprietary
image-generation image-editing google creative visual preview
Released February 2026 · Updated May 16, 2026

Overview

Freshness note: Model capabilities, limits, and pricing can change quickly. This profile is a point-in-time snapshot last verified on May 16, 2026.

Nano Banana 2 is Google’s February 26, 2026 image-generation launch that combines the speed of the Flash line with more of the reasoning and quality previously associated with Nano Banana Pro. Google’s current Gemini API docs now identify Nano Banana 2 as the Gemini 3.1 Flash Image Preview model, with the model ID gemini-3.1-flash-image-preview.

Capabilities

Google’s launch messaging emphasizes four practical upgrades: stronger world knowledge, better text rendering and translation inside images, tighter instruction following, and improved subject consistency. Those are meaningful workflow improvements because they make image generation more useful for infographics, product explainers, localized visuals, and structured visual storytelling, not just aesthetic prompt art.

Google also claims support for maintaining up to five characters and up to 14 objects in a single workflow, plus output sizes from 512px to 4K. For teams building visual assets rather than casual experiments, those are the kinds of details that make the model more operationally relevant.

Technical Details

As an image-generation model, Signal Lens stores contextWindow and maxOutput as 0 and treats them as N/A in the UI. Google’s current API docs present Nano Banana through three distinct lanes:

  • Nano Banana 2: gemini-3.1-flash-image-preview, optimized for speed and high-volume developer use cases.
  • Nano Banana Pro: gemini-3-pro-image-preview, designed for professional assets and complex instructions.
  • Nano Banana: gemini-2.5-flash-image, the older efficient Gemini 2.5 Flash Image lane.

All generated images include SynthID watermarking in Google’s current API docs.

Pricing & Access

Google’s current Gemini pricing docs list the fast native image-generation lane at:

  • Input: $0.30 per 1M text or image tokens
  • Output: about $0.039 per image for outputs up to 1024x1024px

Google’s February 26 launch says Nano Banana 2 is rolling out across Gemini, Search, AI Studio + API, Vertex AI, Flow, and Ads, with preview availability in AI Studio and the Gemini API. For implementation work, use the current API model ID rather than older product shorthand.

Best Use Cases

Use Nano Banana 2 for infographics, product explainers, localized image variants, diagram generation, and high-speed visual iteration where better instruction following matters more than pure art-style experimentation. It is especially relevant for non-creative teams that need useful visuals quickly, not just mood boards.

Comparisons

  • Nano Banana Pro (Google): Higher-fidelity Gemini 3 Pro Image route for professional assets; Nano Banana 2 is the faster iteration lane.
  • Imagen 4 (Google): More established image family route, while Nano Banana 2 is the newer fast-release direction in Google’s stack.
  • Midjourney-style tools: Better for creator-led exploration, while Nano Banana 2 is more relevant to structured product and workflow integration.