Short answer
Nano Banana Pro is Google's Gemini 3 Pro Image model. On TeamoRouter its model ID is gemini-3-pro-image, and you call it through the Gemini native API: POST https://api.teamorouter.vip/v1beta/models/gemini-3-pro-image:generateContent with responseModalities: ["IMAGE"]. Pick the output size with imageConfig.imageSize (1K, 2K or 4K) and the shape with imageConfig.aspectRatio. At Google's list price an image costs about $0.134 at 1K or 2K and about $0.24 at 4K.
What Nano Banana Pro is good at
Google released Nano Banana Pro on November 20, 2025, built on Gemini 3 Pro. Compared with the Flash image models, it reasons about the prompt before drawing, which shows in a few places:
- Text in images. Posters, infographics and menus with legible text, including non-English text.
- Reference images. It can take up to 14 reference images in one request and keep people and products consistent across them.
- High resolution. Native 2K and 4K output, not upscaled 1K.
If you only need fast drafts, the Flash models (Nano Banana 2, gemini-3.1-flash-image) are cheaper; Nano Banana Pro is the one for final images where text and detail matter.
Generate an image
curl https://api.teamorouter.vip/v1beta/models/gemini-3-pro-image:generateContent \
-H "Authorization: Bearer $TEAMO_API_KEY" \
-H "content-type: application/json" \
-d '{
"contents": [
{"role": "user", "parts": [{"text": "A product launch poster for a ceramic mug, headline \"Morning Ritual\", 16:9"}]}
],
"generationConfig": {
"responseModalities": ["IMAGE"],
"imageConfig": {"aspectRatio": "16:9", "imageSize": "2K"}
}
}'
The image comes back as Base64 in candidates[0].content.parts[].inlineData. Loop over all parts to find it rather than assuming it's the first one; the model can return text parts too.
To edit an image, send the instruction and the source image together in parts, with the image as inlineData (mimeType plus Base64 data). The request format is the same as for Nano Banana 2, so the full Python and JavaScript examples in the Gemini 3.1 Flash Image API docs work once you change the model ID to gemini-3-pro-image.
Sizes and aspect ratios
imageSize |
1:1 output | Image output tokens |
|---|---|---|
1K |
1024 × 1024 | 1,120 |
2K |
2048 × 2048 | 1,120 |
4K |
4096 × 4096 | 2,000 |
Aspect ratios: 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9 and 21:9. Write the size in uppercase (2K, not 2k).
Pricing
Google's list price for Nano Banana Pro is $2 per million input tokens and $120 per million image output tokens. With the token counts above, that's about $0.134 per 1K or 2K image and $0.24 per 4K image; 2K costs the same as 1K, so there's little reason to stay at 1K. TeamoRouter bills the same tokens at a discount from the list price (about 70% off in October 2026), with no monthly fee. The current price is on the Gemini 3 Pro Image model page.
For comparison, Nano Banana 2 (gemini-3.1-flash-image) lists at $0.50 per million input tokens and $60 per million image output tokens.
FAQ
Is Nano Banana Pro the same as Gemini 3 Pro Image? Yes. Nano Banana Pro is the product name; Google's API ID is gemini-3-pro-image-preview, and on TeamoRouter it's gemini-3-pro-image.
Can I call it with the OpenAI SDK? Use the Gemini native endpoint above. Image output for Gemini models comes back in Gemini's inlineData format.
Does 4K cost more? Yes, about $0.24 per image at list price against about $0.134 for 1K or 2K.
Can I use the same API key for Claude and GPT? Yes. One TeamoRouter key works for every model, and image generation uses the same balance.