Introduction

Nano Banana is Google DeepMind’s codename for Gemini 2.5 Flash Image, a state-of-the-art model for native image generation and editing. It brings natural-language targeted edits, identity consistency across scenes, multi-image fusion, world-knowledge-guided edits, and SynthID watermarking to keep provenance intact. It’s available in the Gemini app and via API (AI Studio / Vertex AI). Pricing is transparent at about $0.039 per image. :contentReference[oaicite:29]{index=29}

What’s New

Identity Consistency

Keep a person, pet, or product looking like itself across variations—perfect for brand sets or episodic content. :contentReference[oaicite:30]{index=30}

Targeted, Natural-Language Edits

Blur backgrounds, remove objects, change poses, recolor, and more—no masks or layers required. :contentReference[oaicite:31]{index=31}

Multi-Image Fusion

Combine multiple inputs into coherent scenes for product placement, room restyling, or group photos. :contentReference[oaicite:32]{index=32}

Watermarking & Safety

All outputs ship with visible marks and SynthID invisible watermarks for provenance. :contentReference[oaicite:33]{index=33}

Latest Headlines at a Glance

  • Sundar Pichai teased the launch with the “three bananas” post; the codename stuck. :contentReference[oaicite:34]{index=34}
  • LMArena shows the model topping the Image Edit leaderboard. :contentReference[oaicite:35]{index=35}
  • Coverage notes strengths in consistency—and community concerns around deepfakes and basic crop limitations. :contentReference[oaicite:36]{index=36}

Quickstart (API)

Python (google-genai)

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17

from google import genai
from PIL import Image
from io import BytesIO

client = genai.Client()
src = Image.open("input.jpg")
prompt = "Keep identity; swap background to studio lighting."

resp = client.models.generate_content(
    model="gemini-2.5-flash-image-preview",
    contents=[prompt, src],
)
# Save first image part
for c in resp.candidates:
    for p in c.content.parts:
        if getattr(p, "inline_data", None):
            Image.open(BytesIO(p.inline_data.data)).save("output.jpg")

Inline images + text make up a single multimodal request; multiple images enable fusion. See official docs for language guides and Vertex AI parity. (Google AI for Developers, Google Cloud)

TypeScript (@google/genai)

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14

import { Client } from "@google/genai";
import * as fs from "node:fs";

const client = new Client({ apiKey: process.env.GEMINI_API_KEY });

const buf = fs.readFileSync("input.jpg");

const resp = await client.models.generateContent({
  model: "gemini-2.5-flash-image-preview",
  contents: [
    { inline_data: { mime_type: "image/jpeg", data: buf } },
    "Place me in a cozy cafe, keep face and outfit identical.",
  ],
});

Pro Prompt Recipes

  • “Keep face and hairstyle identical; replace background with a narrow Italian alley.”
  • “[Product PNG] + [room photo] → Place product naturally on the shelf; reduce specular highlights.”
  • “[Me], [my dog] → Sunset beach walk; both identities unchanged; realistic lighting.”

Pricing & Access

  • $30 per 1M output tokens (~$0.039 / image), via Developer API and Vertex AI; model is public preview as of Aug 26, 2025. ([Google Developers Blog][3], Google Cloud)

Risks & Responsible Use

Identity consistency is powerful but raises deepfake risks; keep consent, disclosure, and watermark retention in your workflow. (PC Gamer)

Conclusion

Nano Banana isn’t just a new image model—it’s a production-ready editing pipeline for consistent characters and brand visuals. Move fast, test in the app, then wire the API into your toolchain.

Summary

  • Nano Banana = Gemini 2.5 Flash Image (public preview).
  • Strengths: identity consistency, precise local edits, multi-image fusion.
  • Transparent pricing and API parity across AI Studio & Vertex AI.
  • Built-in SynthID watermarking for provenance.
  • Use responsibly; disclose edits and retain watermarks.
1
2
3


**Notice:** :contentReference[oaicite:40]{index=40}
::contentReference[oaicite:41]{index=41}

[3]: https://developers.googleblog.com/en/introducing-gemini-2-5-flash-image/ "

        Introducing Gemini 2.5 Flash Image, our state-of-the-art image model
        
        
        \- Google Developers Blog
        
    "