See how ChatGPT, Perplexity and Google AI Overviews describe you today
Free AI Visibility AuditNano Banana 2 is Google's image generation and editing model, released under the name Gemini 3.1 Flash Image. It takes text and images as input and produces images, aiming for the visual quality of Google's Pro tier at the speed of its Flash tier.
GoogleVerified 24/09/2026Released 18/06/202601 / decision
Each figure sits next to the middle of the field, so it can be read as dear or cheap, wide or narrow, rather than floating on its own. Compared against the 151 models we track, not against an absolute standard.
Capability
Not measured
No published benchmark scores yet
Input / 1M tokens
$0.50
field median $0.43
Output / 1M tokens
$3.00
field median $1.75
Context
131K
field median 524K
02 / overview
Four questions, answered separately, because somebody arrives at one of them rather than at the top of the page.
The job it was built for, and the job it is not for.
Nano Banana 2 is an image model, built to generate pictures from a prompt and to edit pictures you already have. It sits in Google's Flash line, where the trade being made is response speed against the heavier Pro models. It is not a general reasoning or coding model, and it is not built for long-form text output, the text side of it exists to describe and direct the image work.
Where it sits in its line, and when a sibling is the better pick.
Nano Banana 2 first appeared in June 2026 as the image variant of the Gemini 3.1 generation, carrying forward the Nano Banana nickname from the previous release. Pick it when you are generating or editing visuals at volume and want them back quickly, product shots, variations, in-place edits to an existing asset. It is the wrong pick when the job is text, analysis or anything agentic, a general Gemini model handles that instead.
The API, the apps it powers, and what its limits let you do.
You reach Nano Banana 2 through the Gemini API and Google's developer surfaces, and through the Gemini consumer app where image generation and editing sit in the normal chat flow. It accepts both images and text in the same request, so you can hand it a reference picture and describe the change you want rather than re-prompting from scratch. The context window is wide enough to carry several reference images and a detailed brief in one call.
What changes because this exists, or why it does not.
The practical change is that high quality image editing stops being a slow, batched step. Teams that previously queued visual work against a heavier model can now run it inline, generating variants in a UI or regenerating assets as copy changes, without the wait that made it awkward to put in front of a user.
Superseded by Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image). See the whole line.
03 / evidence
Split by category, so a strong number never hides a thin evidence base. Verified means we read it on the benchmark's own published results; a provider's claim about its own model is shown and labelled rather than dropped.
No published benchmark scores for this model yet.
We publish a score only where we can link the result to where it was published. Until a benchmark result for this model exists in a source we read, this section stays empty rather than being filled with a provider's marketing figure.
How we decide what counts as evidence04 / ledger
Every published row, grouped by category, each compared with the best published score on the same benchmark. 'Is 64% good' is a question nobody can answer; '26 points behind the leader' is one anybody can.
Nothing in the ledger yet.
Each row here carries a score, the benchmark it came from, what the leading model scored on the same test, and a link to the published result. Rows appear as results are published and read.
How we decide what counts as evidence05 / capability
Where this model is strong, and against how many peers. Ranks are against models with evidence in that category, not against everything we track: ranking against models nobody tested would rank who published, not who is better.
No category scores to shape yet.
A category score is the weighted mean of the benchmarks published for it. With no published rows there is nothing to average, and an empty chart drawn at zero would say something false.
How we decide what counts as evidence06 / cost
List API rates as last read from the provider, with the source on every row, plus every change we have recorded since we started tracking it.
| Charge | Price | Unit | Read on | Source |
|---|---|---|---|---|
| Input | $0.50 | per 1M tokens | 2026-09-24 | Check |
| Output | $3.00 | per 1M tokens | 2026-09-24 | Check |
Nano Banana 2 is priced at the low end of Google's range, in line with the rest of the Flash family rather than the Pro tier it is being compared to on quality. For most teams the cost is low enough that image generation becomes a routine part of a product flow rather than something metered carefully.
| Workload | Input / month | Output / month | Cost |
|---|---|---|---|
| A small product team | 20M tokens | 5M tokens | $25.00 |
| A busy support assistant | 200M tokens | 40M tokens | $220.00 |
| A document pipeline | 1000M tokens | 100M tokens | $800.00 |
List API rates, no caching and no batch discount, which both providers offer and which change the answer a great deal. Treat these as the ceiling, not the bill.
07 / specs
As listed by the provider's own catalogue and re-read every few hours. Anything absent is absent there too.
| Context window | 131,072 tokens |
|---|---|
| Maximum output | 32,768 tokens |
| Modalities | image, text |
| Released | 18/06/2026 |
| Status | Current |
| Catalogue identifier | google/gemini-3.1-flash-image |
08 / lineage
What this model replaced, what replaced it, and what else its provider has in the field.
Also from Google
09 / line
Every model in this family in release order, so a page from eight months ago says in one glance that two newer ones exist.
Came before
The first we track in this line.
Ordered by release date and worked out from the naming, so a new member slots in as soon as its page exists. A retired model keeps its page and its place in the line.
10 / notes
What this model changes for a brand trying to be cited in AI answers, and every change we have logged since it launched.
What it changes for you
An image model does not cite you, so Nano Banana 2 changes little about how your brand gets named in AI answers. What it does change is the surrounding expectation, buyers arriving at your site have been generating clean visuals in seconds and will judge your product imagery against that. If your category involves visual output at all, expect the question of whether you generate it natively to come up in evaluation.
Most buyers meet Nano Banana 2 inside the Gemini app, where asking for an image or an edit routes to it without the model name ever appearing. Developers meet it directly through the Gemini API and Google's build tooling. It is an image surface rather than an answer surface, so it shapes what a buyer sees more than what they are told.
Nothing published here yet. Changes appear within hours of a provider announcing them.
11 / questions
The things people ask about this model, answered from what is on this page rather than from anywhere else.
Models change what gets cited. We measure whether AI answers name your brand or your competitors across every assistant, and show you what to change.