Gemini 2.5 Flash, widely known as Nano Banana, is currently trending online thanks to its ability to create striking 3D-style models. On social media, users are turning their photos into 3D versions, reminiscent of the Ghibli-style craze that followed ChatGPT’s GPT-4o model a few months ago.
Nano Banana is powerful on its own, capable of generating images from natural-language text prompts and making detailed edits such as adjusting lighting, backgrounds, styles, and textures while keeping the image visually consistent. However, prolonged use may result in reduced image resolution or unintended changes in details. Some users, including myself, find that it sometimes struggles to fully understand complex requests.
This is where AI chatbots like ChatGPT and Qwen can help. They can enhance the outputs from Nano Banana by improving resolution, refining details, and in the case of Qwen, even removing the watermark that appears on the images. Before diving into these improvements, here’s how to create a 3D model of yourself using Gemini.
Steps to Create a 3D Model of Yourself
1. Open the Gemini app.
2. Select Gemini Nano Banana and upload the image you want to turn into a 3D model.
3. Enter a detailed prompt describing the figurine. For example, you could request a one-seventh scale commercial-style figurine placed on a computer desk with a transparent round base, a toy packaging box featuring flat 2D artwork, and a computer screen showing the 3D modelling process.
Once the image is generated, you can use Qwen to enhance the image. Open the Qwen app, select Image Edit, and provide a prompt such as:
“Remove the Gemini logo in the corner and increase the resolution while keeping all other elements intact.”
ChatGPT can also be used to upscale the resolution of Nano Banana’s output, though it cannot remove the watermark. ChatGPT excels in creating precise prompts for Gemini, helping users get better results when the model struggles to interpret their instructions.
Here are some creative prompt examples that ChatGPT can generate for Gemini:
Golden Hour Street Portrait
Transform the photo into a cinematic scene at golden hour on a quiet city street with warm sunlight highlighting the hair, subtle background blur, and soft lens flare.
Desert Sand Statue
Turn the subject into a monumental sandstone statue partially buried in desert dunes with dramatic golden-hour lighting and sand effects.
Ultra-Realistic Oil Painting
Create a photorealistic oil painting of the uploaded image with visible brushstrokes, rich warm tones, soft candlelight shadows, and a detailed ornate frame.
Cyberneon 3D Model
Generate a one-seventh scale action figure with cyberpunk styling, including a matte black jacket with glowing circuit accents, metallic details, and a transparent base. Include a computer screen showing the 3D asset in progress and a premium collector box with neon-lined flat 2D illustrations. Use balanced lighting with cinematic depth of field and subtle film grain for a photorealistic finish.
Streetwear Samurai Collectible
Produce a one-seventh scale collectible that mixes modern streetwear with subtle samurai armour elements, displayed on a transparent base. Include a workspace scene showing the modelling process on a computer screen and a packaging box with elegant flat 2D illustrations printed with a soft-touch cardboard texture. Lighting should create a moody city evening feel with emphasis on fabric and armour details.
By combining Gemini Nano Banana with Qwen and ChatGPT, users can create highly detailed, polished 3D figurines from their photos while maintaining full creative control over style, lighting, and presentation.