Nano Banana Tutorial: How to Use Google Gemini Image Editor

Good news—Google launched Gemini 2.5 Flash Image, a powerful AI image editor and generator, on August 26, 2025. Internally nicknamed “Nano Banana,” this model can create images from prompts, edit your photos with plain-English instructions, blend multiple images, and—crucially—keep characters or subjects consistent across edits. You can try it in the Gemini app and through Vertex AI / the Gemini API. 

Can’t wait to experience this new image model? Let’s explore Nano Banana on Google Gemini—and then see how the Picsman’s Nano Banana makes things even faster with prebuilt scenarios and prompts.

If you’re tracking the landscape of ai image generation in 2025, Nano Banana is one of the biggest shifts: fast, consistent, and widely available.

Table of Contents

1. How to use nano banana on Google Gemini to edit images

✔ Step-by-step guide

1. Open Gemini (web or mobile) and sign in with your Google account.
2. Upload a photo, or start from a text prompt to create your image.
3. Type what you want in plain English, like “Remove the person on the left and blur the background slightly.”
4. Iterate with multi-turn editing: add follow-up prompts like “Now change the jacket to deep navy and add soft studio lighting.” The model remembers context across steps. Learn more in the Google Blog.
5. Export when you’re happy.

✔ Good to know

• Typical edits: remove objects, change backgrounds, blend two photos, recolor, upscale, style-mix, and more.
  
• Respect rights: don’t remove watermarks or edit images you don’t own or have permission to modify.

✔ Core features of the Google Gemini new model

• Blend photos together: Put a person or object from Image [2] into Image [1] with lighting that matches the scene.
  
• Multi-turn editing: The model remembers context so you can tweak step by step.
  
• Style mixing: Transfer style from a reference image to your photo.
  
• Identity consistency: Keep the same face/character across edits and scenes.
  
• API-ready: Works in the Gemini app, AI Studio/API, and Vertex AI—handy for teams.

Not sure how to use the tool or come up with prompts? Try Picsman’s Nano Banana—its categorized toolset makes everything a breeze. Unlike Google Gemini, Nano Banana offers a variety of prebuilt scenarios and prompts.

2. Try Picsman’s Nano Banana: when you want templates and tools

Prefer a guided workflow? Picsman’s nano banana provides pre-built scenarios (Virtual Try-On, Background Replace, Grids, Restoration) and ready-made prompts—great for non-designers. It’s a simple path to Gemini image generation, and users can edit images without writing long prompts.

Firstly, you need to upload your image in Picsman Nano Banana and describe what you want to change in the image, eg change hair color to red. Or you can use our pre-built scenarios directly.

Here are some functions you can use in Picsman’s nano banana.

1) Turn your photo into a figurine

Haven’t met your mini-you yet? Let Nano Banana sculpt you into a collectible.

✅What it does:

Converts a portrait into a realistic “commercial figurine” scene—desk, base, box art and all.

✅Prompts:

• Create a 1/7-scale figurine of this person. Keep the same face and hairstyle. Place on a computer desk with a clear acrylic base, no text.
• Show a ZBrush modeling screen on the monitor and a BANDAI-style box with the original art beside it.
• Make the figurine finish slightly glossy with subtle painted hair highlights and studio top light.

2) Blend two pictures (add a person)

Without a photo with someone? Don’t worry, Picsman’s nano banana can blend two photos. Pop hime or her into the shot—no stress.

✅What it does:

Merges a person from one photo into another while matching light, color, and shadows.

✅Prompts:

• Add the person from Image [2] to the group photo in Image [1], consistent with the environment. Match lighting, perspective, and shadows.
• Match white balance and film grain so the blend looks natural.
• Add soft contact shadows under feet and hands to ground the subject.

Always get consent from everyone pictured.

3) Create a luxury product ad (person + product)

Launching something beautiful? Make it look campaign-ready in seconds.

✅What it does:

Combines a model and product into a high-end poster with fashion lighting and clean details.

✅Prompts:

• Use the person from Image [1] and the product from Image [2] to create a luxury ad. Natural, elegant pose; romantic background; maintain sharp product logos.
• Add a soft reflection under the product and gentle film grain; color palette: champagne, ivory, charcoal.
• Keep proportions realistic; no warping of the product.

4) Stylized studio portrait

Need a pro headshot—fast? Turn a casual snap into studio-grade.

✅What it does:

Cleans skin lightly, sets flattering lights, and preserves identity.

✅Prompts:

• Transform this into a studio-quality portrait: accurate skin tone, soft key at 45°, rim light at 120°, f/2.8 look, gentle bokeh. Keep identity the same.
• Add catchlights in the eyes and a neutral gray background.
• Retouch gently—preserve skin texture (no plastic look).

5) 3×3 or 2×3 photo grid

Can’t pick a look? Try nine at once and choose your favorite.

✅What it does:

Generates multiple styled variations and lays them out in a tidy grid.

✅Prompts:

• Create a 3×3 grid of studio variations: clamshell, Rembrandt, split, butterfly, soft window. Even spacing and thin borders.
• Keep framing and subject size consistent across all tiles.
• Add small captions in the corner naming each lighting style.

6) Edit the text inside your image

Want to edit the text in your image? Fix it like a pro.

✅What it does:

Replaces copy on signs/packaging you own while matching perspective and texture.

✅Prompts:

• Replace the text on this sign with “Welcome to Saturday Market.” Match perspective, font weight, and material texture.
• Use a condensed sans-serif; slight UV-print sheen; keep wood grain visible.
• Warp the text to the board’s angle and keep edge wear realistic.

7) Replace the background or a single object

Same subject, better setting. Move scenes or swap items in one go.

✅What it does:

Places your subject in a new environment or switches one object, then relights to match.

✅Prompts:

• Replace the background with an evening city skyline and shallow depth of field. Keep the subject intact; relight to match the new scene.
• Swap the handbag with the uploaded product while keeping pose and hand position natural.
• Match horizon line, color temperature, and shadow direction; clean up edges around hair.

8) Virtual clothes try-on

New outfit, no dressing room. See styles on your model instantly.

✅What it does:

Applies garments from prompts or images with realistic drape and texture.

✅Prompts:

• Swap the T-shirt for a beige cashmere crewneck; same pose and hair; match sleeve length and natural wrinkles.
• Dress the model in the uploaded red satin evening gown; neutral studio background; realistic fabric sheen and shadow.
• Generate three looks: streetwear, business casual, formal.

9) Enhance or free restore old photos

Have a precious but damaged photo? Bring it back gently.

✅What it does:

Removes scratches and dust, reduces noise, balances tones, and optional colorization.

✅Prompts:

• Restore this 1970s family photo: remove scratches/dust, recover midtones, colorize realistically, keep film grain subtle, preserve faces.
• Light denoise; no over-sharpening; keep the vintage feel.
• Provide a color reference (if available) for more accurate skin tones.

10) Outpaint (expand the image)

Need a wider frame? Grow your canvas without ruining the vibe.

✅What it does:

Extends the scene beyond the original borders, continuing patterns, skies, or floors seamlessly.

✅Prompts:

• For banners, expand more to the left/right and leave clean space for text.
• Extend the image by 30% on all sides. Continue environment seamlessly; maintain perspective, lighting, and textures. Avoid duplicate subjects.
• Add gentle sky gradation, continue the floor tile pattern; keep the subject centered on the rule of thirds.

3. Nano Banana vs. Picsman (comparison table)

Pick the best path for your workflow.

4. Pro tips for better ai images prompt

Make your prompts shorter—and more precise.

• Use a simple formula: Action + Subject + Style/Lighting + Scene/Constraints.
  
• Lock identity: Add “keep same face/identity as original” when editing people.
  
• Constrain the change: Say “only modify background” or “change jacket color only.”
  
• Reference style: Upload a style image for gemini image generation or style-mixing.
  
• Iterate in steps: Small, specific follow-ups > one long prompt.
  
• Save versions: Export intermediate results so you can A/B test.
  

5. Safety, rights, and disclosure

Create responsibly with AI Image Editor tools.

• Edit only content you own or have permission to modify.
  
• Do not remove watermarks from third-party content.
  
• Get consent before altering a person’s likeness.
  
• Disclose AI-generated or AI-altered visuals when required by policy or law.
  

6. FAQs

7. Conclusion

• Pick Google Gemini for flexible, prompt-driven power and API access.
  
• Choose Picsman for fast results with pre-built scenarios and ready prompts.
  
• Either way, nano banana image editing makes it easy to create, refine, and scale visuals for marketing, ecommerce, and social.