AI library

Source · docs

Gemini Nano Banana 2.1 model page

Author
Google
Published
6 October 2026
Added
7 October 2026
Original
https://ai.google.dev/gemini-api/docs/models/gemini-nano-banana-2.1

Summary

Google made Gemini Nano Banana 2.1 (gemini-nano-banana-2.1) generally available on 2026-10-06, according to the Gemini API changelog. It is Google's high-efficiency image generation and conversational image-editing model. It replaces gemini-3.1-flash-image (Nano Banana 2), which is now deprecated with no shutdown date.

Key points

  • Google claims better visual quality, prompt adherence, multi-turn character consistency, and text rendering at "Flash-level" speed and cost.
  • New panoramic aspect ratios: 1:4, 4:1, 1:8, 8:1. Output at 1K, 2K, or 4K for every aspect ratio.
  • Limits: 131,072 input tokens, 32,768 output tokens. Inputs: text, image, video, PDF. Outputs: image and text.
  • Supports thinking (minimal, medium, high), Google web and image search grounding, the Batch API, and up to 14 reference images.
  • No caching, function calling, structured outputs, or Live API.
  • The model page did not list prices when read on 2026-10-07.

Why it matters

This is Google's current image model in the API. Apps that use gemini-3.1-flash-image should plan to migrate.