GeminiGoogle image model

Gemini 3.1 Flash Image. Nano Banana 2, from 512 to 4K.

Google's Nano Banana 2. Make an image from a prompt, or edit and compose from up to 10 reference images on a node, in fourteen aspect ratios at 512, 1K, 2K or 4K.

  • Text and image input
  • 512, 1K, 2K or 4K setting
  • 14 aspect ratios, 1:8 to 8:1
  • Up to 10 reference images

Made with Gemini 3.1 Flash Image

Three different jobs rather than three takes on one: a technical drawing, a photoreal frame and an illustrated scene, each from one prompt at 1K.

A hot air balloon drawn as a technical cutaway in thin ink lines on cream paper, with sage and rust accents
A technical illustration.
A bowl of ramen steaming on a dark wooden counter in moody side light, chopsticks across the rim
A photoreal food frame.
An isometric-style illustration of a rooftop garden apartment with plants, a bicycle and striped awnings
An isometric-style illustration.

Every frame above was generated on Treza withGemini 3.1 Flash Image

Gemini 3.1 Flash Image is Google's image generation and editing model on the Flash tier, also known as Nano Banana 2. One model handles text-to-image and image-to-image: give it a prompt, or a prompt plus up to 10 reference images on a node, and pick a resolution of 512, 1K, 2K or 4K. With no setting it renders 1K.

On Treza it sits in the Image Generation node's picker and in chat, next to Gemini 2.5 Flash Image, the original Nano Banana, and Nano Banana 2.1, its successor. Across 120 production renders at 1K it took about 11 seconds, and 16 at the slowest. OpenRouter bills it per output image token, about $0.067 an image at 1K. It bills from prepaid credits, and a failed generation costs nothing.

512 to 4K

Four resolution settings on the node. A 512 draft bills about two thirds of a 1K image, and the final renders at 2K or 4K from the same prompt.

Up to 10 references

Carry a product, a character and a setting into one image, or attach a photo and describe the change. The original Nano Banana takes 3.

Fourteen aspect ratios

From 1:8 and 9:16 portraits to 16:9, 21:9 and 8:1 banners, plus 4:5 and 5:4 for feeds.

Quick at 1K

About 11 seconds across 120 production renders at 1K, and 16 at the slowest, so it fits inside a chat turn.

What to make with Gemini 3.1 Flash Image

Prompts to try, and what to expect back. Each runs on an Image Generation node with Gemini 3.1 Flash Image picked in the model menu.

“A flat-lay of camping gear, top-down, morning light, 4:5”

A composed product scene at the requested crop.

“Swap the drink in this ad photo for the attached can”

A targeted product swap in an existing asset.

“Our mascot (attached) at a whiteboard presenting a chart”

Your character, carried into a new scene.

“A storyboard frame of a kitchen at night, a kid reaching for the cookie jar”

A 512 draft to check the framing before the full-size render.

“The approved draft (attached) at 4K, same composition”

The chosen frame, finished at full size.

“Four thumbnail variants of the same scene, different colorways”

A quick batch to A/B against each other.

Gemini 3.1 Flash Image is one node on a canvas

Chain it with language, image, and video models, then publish the whole pipeline as an API. See how a full AI video pipeline fits together.

  1. Step 01

    Draft the brief

    A language model node expands a one-line idea into a detailed, on-brand prompt before it ever reaches the model.

  2. Step 02

    Generate with Gemini 3.1 Flash Image

    Run variants in parallel, compare against other models on the same prompt, and keep every result in run history.

  3. Step 03

    Publish it as an API

    The pipeline becomes a versioned endpoint. Your product calls it with the typed /invoke API or any OpenAI SDK.

Gemini 3.1 Flash Image as an API, priced per image

The Gemini 3.1 Flash Image API

Build a pipeline with Gemini 3.1 Flash Image on a node and publish it. It becomes a versioned HTTP endpoint: the typed /invoke route enqueues a run and returns a run id to poll, and an OpenAI-compatible route works with SDKs you already use. Agents connect over MCP the same way. Your prompt engineering ships inside the pipeline, so every caller gets your quality, and swapping the model later does not change the API your product calls.

Gemini 3.1 Flash Image pricing

Prepaid credits, no subscription. Each image is charged at Gemini 3.1 Flash Image's own rate from your balance, shown before you run, and a failed generation costs nothing. Credit packs start at $5 and never expire, which suits spiky production schedules. Run history records the cost of every generation, per node, so spend is auditable rather than a surprise.

For agents with a wallet: Gemini 3.1 Flash Image is also sold per call over x402, paid in USDC with no account or key. Pay-per-call images ->

Gemini 3.1 Flash Image, answered

Is Gemini 3.1 Flash Image the same as Nano Banana 2?

Yes. Nano Banana 2 is Google's name for Gemini 3.1 Flash Image. On Treza it is listed in the model picker as Nano Banana 2 (Gemini 3.1 Flash Image), and chat knows it by either name.

How is it different from Gemini 2.5 Flash Image and Nano Banana 2.1?

Gemini 2.5 Flash Image, the original Nano Banana, has no resolution setting and takes 3 references. 3.1 adds 512 to 4K, four more aspect ratios and up to 10 references, at a higher price per image. Nano Banana 2.1 is its successor: the same fourteen aspect ratios and 10 references at 1K, 2K or 4K, for about half the price per image at 1K and 2K. All three run in chat and on the same node, so try one prompt on each.

What does Gemini 3.1 Flash Image cost per image?

OpenRouter bills it per output image token, which comes to about $0.045 an image at 512, $0.067 at 1K, $0.10 at 2K and $0.15 at 4K. Treza quotes those amounts, plus its credit markup, before a run, and a failed generation costs nothing.

Can I use it in chat?

Yes. Ask for Gemini 3.1 Flash Image or Nano Banana 2 by name in chat, attach reference images if you have them, and the image comes back in the conversation at 1K. For 512, 2K or 4K, set the resolution on an Image Generation node.

How much does Gemini 3.1 Flash Image cost on Treza?

Treza uses prepaid credits with no subscription. Each image is charged at Gemini 3.1 Flash Image's rate from your balance, and only successful generations are charged. Credit packs start at $5 and never expire.

Can I call Gemini 3.1 Flash Image through an API?

Yes. Build a pipeline that uses Gemini 3.1 Flash Image and publish it as a versioned HTTP endpoint. Call the typed /invoke API from your product, or point any OpenAI SDK at the OpenAI-compatible endpoint. If a better model ships later, swap it on the node without changing the API your product calls.