Nano Banana 2
Nano Banana 2 provides high-quality image generation and conversational editing at a mainstream price point and low latency.
Modalities
In / Out Price
30% off$0.35/ 1M
$0.5 Official price
$2.1/ 1M
$3 Official price
Context
131.072K
Released
May 28, 2026
Playground
Output
Explore AI StudioAPI call examples
POST /v1/chat/completionscurl --request POST 'https://api.tokatlas.com/v1/chat/completions' \
--header "Authorization: Bearer $UNIROUTER_API_KEY" \
--header 'Content-Type: application/json' \
--data '{
"model": "gemini-3.1-flash-image",
"messages": [
{
"role": "user",
"content": "Explain quantum entanglement."
}
],
"temperature": 0.7
}'Providers & routes
Current provider · Google
Platform configured price · Group: default
USD / 1M tokens. Each row shows input and output from the same tier. Displayed rates use the lowest available group multiplier; your API key group may differ.
Gemini 3.1 Flash Image for reusable asset variants
Use it to draft compositions and refine details through references and follow-up edits. Lock the main subject’s shape and position for product work.
Text-driven drafts
Describe subject, camera angle, and whitespace for a consistent first pass.
Reference-based editing
Use an approved image to guide shape and placement while changing seasonal props.
Consistent framing
Use one aspect ratio so variants can sit side by side.
Delivery checks
Verify brand features and copy-safe areas before cropping for delivery.
About
Where Nano Banana 2 fits
Gemini 3.1 Flash Image is designed for low-latency, high-volume generation and conversational edits. It is a practical option for large draft sets, social artwork, and localized variants alongside the Pro tier.
Sizes, aspects, and inputs
Google lists 0.5K, 1K, 2K, and 4K with 1K as the default, plus formats such as 1:4, 4:1, 1:8, and 8:1. Native inputs include text, image, video, and PDF, with image and text outputs. Confirm the assets and sizes on your actual route.
Grounding and lettering
The upstream model can use text and image search to inform generation when the endpoint and settings expose it. Test a fixed set of package labels and localized copy across languages, sizes, and aspect ratios.
Use smaller images to approve composition and palette, then inspect lettering and edges at higher resolution. For unusually wide or tall assets, verify subject placement after cropping instead of treating one prompt as equally suitable for every ratio.
Nano Banana 2 FAQ
How should I pick a Nano Banana 2 output size?
Use a smaller output to check composition and a larger one for final assets. Pick from the available sizes and check that a new aspect ratio does not crop the subject.
How should I prompt for a very tall or wide layout?
Select an aspect ratio close to the final asset, then specify subject placement and room for copy. Check readability at the actual placement size.
Can I reuse the same composition for a thumbnail and a campaign hero?
If several sizes are available, start small to check the focal point, then render for the final placement. Review the crop at each size.
Does the endpoint provide both text analysis and image output?
Check that you are using the image generation or editing capability listed here. For structured extraction, select a route that explicitly returns text.