What is Nano Banana Pro?
Nano Banana Pro is the nickname for Google's Gemini 3 Pro Image model. It creates images from text and, more importantly, edits photos you upload while keeping the people in them recognisable. That identity preservation is why almost every viral AI photo trend of the past year — saree edits, 90s retro portraits, 3D figurines, scrapbook collages — runs on it.
Step 1: Open Gemini and pick image creation
Go to gemini.google.com or open the Gemini app and sign in with a Google account. In a new chat, choose the image option ("Create image") or simply ask for an image — Gemini routes it to Nano Banana automatically. Free accounts get a daily allowance; Google AI Pro raises it.
Step 2: Attach your photo (for edits)
Tap the + button and upload one clear photo. The best results come from a sharp, front-facing photo in daylight with your full face visible — no sunglasses, heavy filters or group shots unless the prompt is for a couple or group. For outfit or figurine prompts, use a full-body photo.
Step 3: Paste the prompt and send
Copy a prompt from Hixa with the Copy button and paste it into the same message as your photo. Don't trim it — the long details about lighting, camera and face preservation are what make the result look real. If the prompt mentions an aspect ratio, keep it.
Step 4: Refine with short follow-ups
Instead of rewriting the prompt, reply in the same chat: "make the light warmer", "change the saree to emerald green", "crop to 9:16", "remove the text". Gemini keeps the previous image in context, so small, single-change follow-ups give the most control.
Fix: my face looks different
Upload a higher-resolution photo, make sure the face is well lit and unobstructed, and reply: "Keep my face exactly the same as the uploaded photo — same facial features, skin tone and expression." Very stylised prompts (anime, 3D) will always simplify features a little.
Fix: the result ignores part of the prompt
Long prompts sometimes drop a detail. Ask for the missing element on its own in a follow-up — "add the polaroid frame in the bottom-left" — rather than regenerating everything.






