What is Nano Banana (I2I)?

Nano Banana (I2I) is an AI model (image to image) available on Artifio.ai through google. You can run it in our workspace from a single shared wallet and download your outputs.

How much does Nano Banana (I2I) cost on Artifio?

Nano Banana (I2I) starts at about $0.020 per generation on Artifio.ai. The exact cost is shown in the workspace before you confirm and can vary with the options you select.

How do I use Nano Banana (I2I) on Artifio?

Click Generate Now to open the workspace at /create/google-nano-banana-edit. Fill in the form fields, review the live cost estimate, and run your generation.

What kind of outputs does Nano Banana (I2I) produce?

Nano Banana (I2I) is configured for image to image workflows in the image category. See the technical parameters below for supported inputs and options.

Can I try Nano Banana (I2I) for free?

Sign up free, top up your wallet from $5, and you can run Nano Banana (I2I) pay-as-you-go without a separate subscription for this model.

Nano Banana (I2I) — AI Image Generator

Name: Nano Banana (I2I)
Brand: Artifio
Price: 0.020 USD
Availability: InStock

$1 on us$1 on us + a free Gold trial to get you startedClaim your $1 + trial Terms apply

artifio.ai

Nano Banana (I2I) — AI Image Generator | Artifio.ai

About Nano Banana (I2I)

Nano Banana is the community nickname for Gemini 2.5 Flash Image, announced by Google on August 26, 2025. Before the launch it circulated anonymously on the LMArena evaluation platform as "nano-banana", where its photo edits went viral and drew over 2.5 million votes; editing, more than generation, is what made the name stick.

This page covers the image-to-image mode: you supply one or more photos and describe the change in plain language. On artifio.ai it sits alongside 100+ other models in one workspace, paid per generation rather than per subscription.

Features

Performs targeted local edits from natural-language instructions: Google's launch examples include blurring a background, removing a stain from a t-shirt, removing a person, changing a subject's pose, and colorizing a black-and-white photo.
Supports multi-turn conversational editing, so one image can be refined step by step across a chat-style session.
Blends multiple input photos into one output while preserving detail from each, such as merging a person, a pet, and a new setting.
Maintains facial and subject likeness through edits, the consistency property that drove the model's viral spread.
Uses Gemini's world knowledge during edits, letting instructions reference real objects and styles by name.
Embeds an invisible SynthID watermark in every edited output.

What people use it for

Retouching personal photos with plain-language requests instead of manual masking.
Placing a product into new backgrounds and scenes while keeping the product itself unchanged.
Restoring or colorizing old photographs.
Combining reference images, such as furniture photos plus a room shot and a color palette, into a single rendered concept.
Iterating on a marketing visual over several conversational turns until it matches the brief.

Reported strengths

Ranked first on the LMArena image-edit leaderboard at launch with a 171-point Elo lead, which LMArena described as the largest in Arena history.
Face and subject consistency through edits was widely cited as the model's standout trait, holding likeness where earlier editors drifted.
Low latency makes conversational, multi-round editing practical rather than a batch process.
Handles compositional edits that need real-world understanding, like relighting a scene plausibly or matching perspective when merging photos.

Reported limitations

Edited outputs are capped at 1024x1024 pixels, so it cannot deliver the 2K or 4K files that Nano Banana Pro and Nano Banana 2 later added.
Google acknowledged at launch that character consistency is not fully reliable and that fine factual details in images still miss.
Safety rules prohibit certain edits, including non-consensual intimate imagery, and consumer surfaces add a visible watermark on top of SynthID.
Long or dense text inside images was a known weak point of this generation of the model, per Google's own launch post.

Sources

Introducing Gemini 2.5 Flash Image, our state-of-the-art image model (Google Developers Blog)
Google Gemini's AI image model gets a 'bananas' upgrade (TechCrunch)
Nano-Banana (Gemini 2.5 Flash Image): Try it on LMArena (LMArena)
Gemini 2.5 Flash Image (Nano Banana) model card (Google AI for Developers)