Google/nano-banana-2-1/Released Oct 2026

Nano Banana 2.1

Fast 4K image generation and editing

Commercial use

Image

About Nano Banana 2.1

Nano Banana 2.1 is Google's Flash-tier image generation and editing model, an update to Nano Banana 2 with better visual quality, prompt adherence, text rendering, and character consistency across edits.

It generates from a text prompt alone, or edits and fuses up to 14 reference images, keeping up to 4 characters and 10 objects consistent. Output comes at 1K, 2K, or 4K across standard and panoramic aspect ratios (up to 8:1 and 1:8), with fewer tiling artifacts on wide formats at 2K and 4K than Nano Banana 2. Images can be grounded in Google Web Search and Google Image Search for factual or current subjects, and a configurable thinking level trades latency for better handling of complex compositions.

See what this model can do

A handful of real prompts and their outputs.

A bilingual Thai and English menu sign above a Bangkok street-food cart at night

Multilingual text

Street Food Menu

"Hand-painted Thai script and English with baht prices, rendered in a phone-photo street scene."

Environmental portrait

Knife Sharpener

"Film-grain candid portrait with mixed bulb and daylight, water beading on the steel."

Exploded-view drawing of a manual burr coffee grinder with a numbered parts list

Technical illustration

Exploded Grinder

"Seven parts on one axis, numbered callouts, and a parts table that matches them."

Golden-hour aerial photo of the terraced Maras salt ponds in the Peruvian Andes

Image Search grounding

Maras Salt Terraces

"Grounded with Image Search for the real terraces, valley, and Andes behind them."

False-colored scanning electron micrograph of four pollen grains with a 10 µm scale bar

Scientific imaging

Pollen Micrograph

"Four distinct pollen types in SEM detail, false-colored, with a scale bar."

Packaging design

Tea Box Dieline

"Magenta cut lines, cyan folds, a glue flap, and print-ready panel artwork."

A museum specimen drawer of twelve pinned beetles, each with a typed Latin-name label

Scientific specimens

Beetle Drawer

"Twelve distinct species, each matched to a typed Latin label, in an aged museum drawer."

Wordless flat-pack assembly instructions for a wooden birdhouse in six steps

Instruction manual

Birdhouse Assembly

"Six numbered steps, a hardware inventory, and no words, in flat-pack manual style."

An architectural floor plan of a one-bedroom apartment with labeled rooms, dimensions, furniture, a north arrow and a scale bar

Architectural drawing

Apartment Floor Plan

"Room labels and dimensions, door swings, furniture, a scale bar and a title block, all legible."

Ready to build with Nano Banana 2.1?

Try Nano Banana 2.1 in the Workbench to prompt it, compare outputs, and iterate on prompts without writing any code. When you're ready to ship, call the same model from our API and build your own apps on top of it.

curl -sSf -X POST https://hub.oxen.ai/api/ai/images/edit \
    -H "Content-Type: application/json" \
    -H "Authorization: Bearer $OXEN_API_KEY" \
    -d '{
  "model": "nano-banana-2-1",
  "prompt": "A vintage travel poster of an ox crossing a snowy mountain pass at sunrise, bold art-deco lettering across the top reading \"VISIT THE HIGH PASS\". Warm oranges against cool blues.",
  "aspect_ratio": "16:9",
  "resolution": "2K",
  "google_search": false,
  "image_search": false,
  "thinking_level": "medium"
}'

Request parameters

Every field Nano Banana 2.1 accepts in the request body. See the API reference for response formats.

POSThttps://hub.oxen.ai/api/ai/images/edit

Field	Type	Required	Default	Description
`prompt`	`string`	Yes	—	Text description of what you want to generate, or the instruction on how to edit the given image. Use @Image1, @Image2, etc. to reference the input images.
`input_image`	`array<string>` nullable	No	—	Optional reference image(s) to edit or combine, up to 14. Leave empty to generate from the prompt alone. Reference them in the prompt as @Image1, @Image2, etc.
`aspect_ratio`	`string`	No	`"16:9"`	`auto1:116:99:164:33:43:22:34:55:421:94:11:48:11:8`
`resolution`	`string`	No	`"2K"`	`1K2K4K`
`google_search`	`boolean`	No	`false`	Use live web results for current facts.
`image_search`	`boolean`	No	`false`	Use Google images as visual references.
`thinking_level`	`string`	No	`"medium"`	Higher levels handle complex scenes but run slower. `minimalmediumhigh`

Model JSON Image editing reference Async queue Models endpoint