How can I access Nano Bananas?

‍‍Discover Nano Banana — Google’s codename for the Gemini 2.5 Flash Image model. Learn how this AI tool edits images via natural language and why it matters.

TL;DR

• Nano Banana is Google’s codename for Gemini 2.5 Flash Image, a fast, high-quality image generation and natural-language editing model. It blends images, keeps characters consistent across scenes, and embeds SynthID watermarking for transparency.
• You can use it in the Gemini app, Promptus, AI Studio / Gemini API, and Vertex AI.
• The most straightforward, all-in-one path for most creators is Promptus: select the Nano Banana / Gemini 2.5 Flash Image model, make edits in plain language, save to Collections, publish to a public gallery, and even turn images into videos and add music—all in one place.

What “Nano Banana” actually is (and why you keep seeing bananas) 🍌

“Nano Banana” is the internal nickname Google uses for Gemini 2.5 Flash Image, a new image model that focuses on speed and control. It supports multi-image fusion, character consistency, targeted, natural-language edits, and leverages Gemini’s world knowledge to understand scenes and instructions. All outputs carry an invisible SynthID watermark (and, in some surfaces, a visible watermark) for responsible, tamper-resilient disclosure.

• Google’s developer announcement explicitly calls it “aka nano-banana” and details the feature set.
• The official model page notes it was tested publicly under the “nano-banana” name before launch.
• In the Gemini app, Google emphasizes visible and digital watermarking for AI-generated or AI-edited media.

Why Promptus is the simplest way for most people

Plenty of paths let you try Nano Banana, but Promptus stands out for one reason: it’s one web interface where you can generate and edit with Nano Banana, organize your results, publish them as a gallery, and even expand your visuals into video and music—without juggling multiple tools or exports.

Here’s what that looks like in practice:

• Model access: Promptus features a Gemini Flash 2.5 / Nano Banana model page and selector, so you can pick the Google model inside the same workspace you use for all your other creative tasks.
• Create → Save → Publish, in one place: Use Collections to store your images and videos; share or publish to a public gallery right from the platform.
• Move beyond stills: Promptus markets built-in tools for AI video and music/audio creation, letting you animate an image into a video and generate soundtrack elements without leaving the site.

In short, if you want the fewest moving parts—access to Google’s Nano Banana model plus saving, organizing, publishing, and multimedia add-ons—Promptus is a very strong, one-stop choice.

Plain-English answer to “How can I access nano bananas?”
Open Promptus, choose Gemini Flash 2.5 / Nano Banana from the model picker, upload (or start from text), describe your edits in natural language, and generate. Save to Collections, publish to your gallery, and—if you want—turn it into a video and add music from the same dashboard.

Step-by-step: Use Nano Banana in Promptus (no technical setup)

1. Open Promptus and sign in.
2. Pick the model: In the model dropdown, choose Gemini Flash 2.5 (the interface may label it “Nano Banana” or similar).
3. Start with an image or a prompt:
   • Edit an image: Upload one or more photos.
   • Generate from scratch: Type what you want to see.
4. Describe your changes in everyday language (examples below). Nano Banana is built for targeted, natural-language edits.
5. Iterate: Ask for refinements (“slightly warmer light,” “retain fabric texture,” “keep the same person but change the outfit”).
6. Save to Collections (your personal library), then publish directly to a public gallery if you’d like to share or sell your work.
7. Go multimedia: From your project, animate into a video and add music/audio using Promptus’ built-in tools—no extra apps needed.

What to try first (copy-paste prompts)

Nano Banana responds well to clear, local instructions and multi-step edits:

• Cleanup & polish
  • “Remove the coffee stain on the shirt but keep the cloth texture unchanged.”
  • “Reduce glare on the glasses without changing the eyes.”
  • “Increase sharpness on the subject; keep the background soft.”
• Character consistency
  • “Keep this person’s face/features the same across all edits. Generate three scenes: office desk, café, beach boardwalk. Vary outfits but preserve hairstyle and glasses.”
• Multi-image fusion
  • “Take the red mug from image A and place it on the desk in image B near the keyboard; match lighting and color temperature.”
• Recolor & relight
  • “Recolor the jacket to forest green and add subtle warm sunset lighting with soft rim light.”
• World-knowledge edits
  • “Use this hand-drawn floor plan to produce a clean colored top-down layout with labels for kitchen, living room, and balcony.”

Each of these plays to the model’s strengths—local, targeted edits; cross-image reasoning; and scene-aware transformations—outlined in Google’s launch materials.

Publishing, galleries, and going beyond still images

One major reason Promptus feels “simple” is what happens after you click Generate:

• Collections & galleries: Curate private Collections or publish to a public gallery on your own Promptus site; Promptus’ docs highlight sharing and even selling images or videos from these galleries. If you’re a creator, that’s your distribution loop solved.
• Video & music in the same place: Promptus’ product pages emphasize AI video and music/audio features. That means you can animate an image into a short clip and add soundtrack elements within the same workspace—no export/import dance.

Quality, speed, and watermarking

• Speed & control: Nano Banana (Gemini 2.5 Flash Image) is designed for low latency while maintaining fine-grained, natural-language edits and character consistency.
• Watermarks: Google says all images created or edited with this model include SynthID (invisible) for provenance, and the Gemini app also applies a visible mark. If you’re publishing widely, this helps with responsible disclosure.

Cost notes (for developers who care)

For API-based work, Google’s pricing post lists $30 per 1M output tokens and roughly 1290 output tokens per image (≈ $0.039 per image on the output side; inputs follow Gemini 2.5 Flash pricing). If you’re scaling, always confirm current rates in Google’s docs.

Alternatives (and when you might use them)

While this guide puts Promptus first for simplicity, Nano Banana is also available here:

• Gemini app (web & mobile) — quick, consumer-friendly edits with watermarking. Great for one-offs and social posts.
• Google AI Studio / Gemini API — the developer playground and programmatic path to production. Ideal for integrating into your product or pipeline.
• Vertex AI — enterprise deployment with governance and MLOps. If you need IAM, quotas, and auditability by default, start here.

Still, if your goal is “use Nano Banana with the least friction and publish what I make,” Promptus remains the clearest route because it unifies creation, asset management, and publishing—and extends your stills into video and music without extra tools.

Troubleshooting: quick answers

I can’t find the Nano Banana model in Promptus.
Look for “Gemini Flash 2.5” (Promptus’ model page uses that name) or similar wording; some UIs label it with the official name first and the codename second.

Do I need to learn complicated parameters?
Not to get started. Nano Banana is built for plain-English instructions and local edits, so you can simply describe what you want changed.

Are there usage rules for watermarking?
Yes. Google embeds SynthID invisibly and, in some surfaces, a visible mark. If you’re publishing or selling work, keep disclosures intact and follow platform guidelines.

A simple starter workflow (bookmark this)

1. Open Promptus→ Select “Gemini Flash 2.5 / Nano Banana.”‍
2. Upload a photo (or start from text).
3. Type one clear instruction (e.g., “Remove the stain on the jacket; keep fabric texture”).
4. Iterate with small changes (“slightly warmer light,” “add a soft vignette”).
5. Save to Collections → Publish to a public gallery (optional).
6. Animate to video → Add music if you want a share-ready clip.

That’s it—from first prompt to public post, without leaving one browser tab.

Final thought

Nano Banana (Gemini 2.5 Flash Image) brings together speed, control, and world-aware editing in a way that makes everyday creators feel like VFX pros. And while there are multiple legitimate access points, Promptus makes the experience obvious and easy: pick the model, create, save, publish, and extend to video + music—all in one place. If your priority is to get results fast and share them, this is the path that removes the most friction.