Nano Banana 2
Nano Banana 2 is the consumer name for Gemini 3.1 Flash Image, the image model Google DeepMind released on 26 February 2026 and made the default across the Gemini app, Search's AI Mode, Lens, Google Ads and Flow. It is the generalist of the family: most of Nano Banana Pro's quality at Flash speed, about half the price, and the one free Gemini accounts actually use.
Anvisha Pai, Co-founder & CEO, ModaTested Google DeepMind · released
›What we asked for
A self-portrait of Nano Banana 2, Google's fast everyday AI image model, as a bold flat vector illustration: a cheerful banana character in motion, leaning forward mid-stride with speed lines, limited palette of yellow, teal and off-white, thick clean shapes, no outlines, poster composition.
fal-ai/nano-banana-2
aspect_ratio: 4:3, resolution: 1K, output_format: png, num_images: 1
returned 1200 by 896 pixels
At a glance
- Sizes
- Four sizes, 512px, 1K, 2K and 4K, in fifteen fixed shapes including extreme 8:1 strips.
- Thinking
- Off unless you ask. The high setting makes it plan before drawing for a small surcharge; the ladder below shows what that buys.
- Reference photos
- Up to 10 object references, 4 character references and 3 style references in one request; consistency for up to 5 people.
- Web grounding
- Can search the web and Google Images before drawing, the only Nano Banana tier with image search.
- No transparent backgrounds
- Output is always opaque.
Other Nano Banana models: Nano Banana 2 Lite, Nano Banana Pro. Compare all 3.
Anvisha's take
This is the model I would put in front of a new user. It was as accurate on set text as Pro on our run, every picture came back in nine to fourteen seconds, and it costs half as much. The habits to watch are the ones it shares with the family: an unrequested 'Learn More' button on a banner, an embossed brand name on an edited product. Say 'no text' when you mean it and it behaves.
Nano Banana 2 real output: the same 12 prompts every model gets
Every model on this site gets the same 12 prompts, so you can compare like with like. These are the pictures Nano Banana 2 returned, one try each, nothing re-rolled or retouched. Generated on . The 12 pictures took 9 to 14 seconds each and cost $0.96 in total.
Spelling, kerning and hierarchy of three lines of set text.
›What we asked for
A concert poster that reads 'MODA SUMMER SESSIONS' in large bold type, with the subtitle 'Friday 12 June, 8pm, Roundhouse' and a smaller line 'Tickets at the door'. Risograph style, two colours, tangerine and navy, grainy ink texture.
fal-ai/nano-banana-2
aspect_ratio: 3:4, resolution: 1K, output_format: png, num_images: 1
returned 896 by 1200 pixels
Catalogue lighting, a clean seamless background and a believable contact shadow.
›What we asked for
Studio product photograph of a matte black ceramic pour-over coffee dripper on a seamless white background, soft top light, subtle contact shadow, centred, catalogue style.
fal-ai/nano-banana-2
aspect_ratio: 1:1, resolution: 1K, output_format: png, num_images: 1
returned 1024 by 1024 pixels
A natural face, correct finger count and hands holding an object.
›What we asked for
Close-up portrait of a woman in her 40s with short grey hair, laughing, holding a red ceramic mug with both hands, natural window light, 85mm lens, shallow depth of field.
fal-ai/nano-banana-2
aspect_ratio: 3:4, resolution: 1K, output_format: png, num_images: 1
returned 896 by 1200 pixels
A flat single-colour mark with a legible wordmark and no gradients or photo texture.
›What we asked for
A flat vector logo mark for a bicycle repair shop called 'Spoke & Chain': a single-colour icon combining a chain link and a wheel spoke, with the wordmark below in a geometric sans-serif, black on white.
fal-ai/nano-banana-2
aspect_ratio: 1:1, resolution: 1K, output_format: png, num_images: 1
returned 1024 by 1024 pixels
Many distinct objects and people in one frame without merging or duplicating them.
›What we asked for
A crowded Saturday farmers market seen from slightly above: at least twelve stalls with striped awnings, shoppers carrying tote bags, a dog on a leash, crates of oranges, a musician with a guitar, bunting overhead, late morning light.
fal-ai/nano-banana-2
aspect_ratio: 4:3, resolution: 1K, output_format: png, num_images: 1
returned 1200 by 896 pixels
A named historical style applied faithfully rather than a generic painterly look.
›What we asked for
A lighthouse on a rocky cliff during a storm, in the style of a Japanese ukiyo-e woodblock print, with visible wood grain texture, flat colour areas and a bokashi gradient sky.
fal-ai/nano-banana-2
aspect_ratio: 3:4, resolution: 1K, output_format: png, num_images: 1
returned 896 by 1200 pixels
Six labelled bars at the stated heights, in order, with readable axes.
›What we asked for
A clean bar chart titled 'Monthly active users' showing six months, January through June, with the values 120, 150, 180, 170, 220 and 260, labelled axes, white background, one accent colour, no 3D effects.
fal-ai/nano-banana-2
aspect_ratio: 4:3, resolution: 1K, output_format: png, num_images: 1
returned 1200 by 896 pixels
A 21:9 composition that leaves the left third clear for headline text.
›What we asked for
Website hero banner: a wide panorama of a sunlit modern co-working space with plants, the left third of the frame kept as clear wall space for text, warm neutral palette.
fal-ai/nano-banana-2
aspect_ratio: 21:9, resolution: 1K, output_format: png, num_images: 1
returned 1584 by 672 pixels
Four numbered steps in the given order with their exact wording, plus an icon each.
›What we asked for
An infographic card titled 'How to brew pour-over' with four numbered steps in this order: 1. Rinse the filter, 2. Bloom for 30 seconds, 3. Pour in slow circles, 4. Wait 3 minutes. A simple icon beside each step, muted palette, clean sans-serif type.
fal-ai/nano-banana-2
aspect_ratio: 3:4, resolution: 1K, output_format: png, num_images: 1
returned 896 by 1200 pixels
Appetising, menu-grade food with correct textures and no melted anatomy.
›What we asked for
Overhead photograph of a margherita pizza on a wooden board, fresh basil leaves, one slice pulled out with the cheese stretching, restaurant menu photography, natural light.
fal-ai/nano-banana-2
aspect_ratio: 1:1, resolution: 1K, output_format: png, num_images: 1
returned 1024 by 1024 pixels
Exact counts and left-right placement of five named objects.
›What we asked for
A glass of water on a wooden table with exactly three ice cubes in it, a yellow pencil resting across the rim of the glass, a blue book to the left of the glass and a red apple to the right, photographed from table height.
fal-ai/nano-banana-2
aspect_ratio: 4:3, resolution: 1K, output_format: png, num_images: 1
returned 1200 by 896 pixels
Change one object's colour and add a part while leaving background, light and framing untouched.
›What we asked for
Change the coffee dripper to a deep forest green and add a small brass handle on its right side. Keep the background, lighting and composition exactly as they are. (applied to the product on white result above)
fal-ai/nano-banana-2/edit
aspect_ratio: 1:1, resolution: 1K, output_format: png, num_images: 1
returned 1024 by 1024 pixels
How the cost under each picture was worked out: fal unit price, 0.08 USD per image, read from fal's pricing API at run time.
Nano Banana 2 settings that matter
Thinking level
High thinking costs a fifth of a cent and a couple of seconds. It did not make either picture sharper; it changed what was drawn. On the poster, minimal invented an extra address line at the bottom while high kept to the three lines asked for and set them bolder. On the pizza both added a menu card with garbled small print. Turn it on for layouts and text, leave it off for photographs.
Text in an image
›What we asked for
A concert poster that reads 'MODA SUMMER SESSIONS' in large bold type, with the subtitle 'Friday 12 June, 8pm, Roundhouse' and a smaller line 'Tickets at the door'. Risograph style, two colours, tangerine and navy, grainy ink texture.
fal-ai/nano-banana-2
aspect_ratio: 3:4, resolution: 1K, thinking_level: minimal, output_format: png, num_images: 1
returned 896 by 1200 pixels
Food photography
›What we asked for
Overhead photograph of a margherita pizza on a wooden board, fresh basil leaves, one slice pulled out with the cheese stretching, restaurant menu photography, natural light.
fal-ai/nano-banana-2
aspect_ratio: 1:1, resolution: 1K, thinking_level: minimal, output_format: png, num_images: 1
returned 1024 by 1024 pixels
Nano Banana 2 pricing
| Where | Unit | Price | Source |
|---|---|---|---|
| Gemini API | 1K image $60 per million image output tokens. Google's own per-image figures: $0.045 at 512px, $0.067 at 1K, $0.101 at 2K, $0.151 at 4K. | $0.067 | Gemini API pricing, read |
| Gemini API | 4K image | $0.151 | Gemini API pricing, read |
| Gemini API, Batch | 1K image Batch halves every rate. | $0.034 | Gemini API pricing, read |
| fal | 1K image Flat per-image price on the generate and edit endpoints, the basis for our sample run. 512px is $0.06, 2K $0.12, 4K $0.16. High thinking adds $0.002 and web search $0.015. | $0.08 | fal model page and pricing API, read |
| fal | 4K image | $0.16 | fal model page, read |
Is Nano Banana 2 free?
This is the free one. Free Gemini accounts generate with Nano Banana 2 (or Nano Banana 2 Lite depending on the model setting) and download at 1K. The Gemini API has no free tier for it.
- •Gemini app, free: image generation runs on Nano Banana 2 at 1K. Google publishes no daily number and says limits may change without notice, including for capacity.
- •Gemini app, Google AI Plus, Pro and Ultra: 2K downloads and the option to redo an image with Nano Banana Pro.
- •Gemini API: the free tier is listed as not available for gemini-3.1-flash-image.
- •Every image carries an invisible SynthID watermark.
Where to use Nano Banana 2
- Gemini app · The default image model, free plan included.
- Google AI Studio and Gemini API · Model id gemini-3.1-flash-image. Pay per token.
- fal · Flat $0.08 per 1K image, generate and edit endpoints.
- Replicate · Hosted endpoint under google/nano-banana-2.
- Moda · Available as the NanoBanana 2 image model inside the Moda design agent, with 512px to 4K output and the thinking control.
Generate with Nano Banana 2 in Moda
Pick Nano Banana 2 in the Moda design agent, then edit the result as vectors, text and layers rather than a flat image.
Open Moda →What we noticed running Nano Banana 2
- •Set text is as reliable as Pro's. All three poster lines came back correct on the first try, with a two-colour illustration of a round venue and a crowd added underneath. See the text in an image picture.
- •Edits can leak real brands. The recolour and brass handle arrived as asked, and so did an embossed 'V60' on the base, a trademark of a real pour-over maker that was never mentioned. Check edited product shots for logos before they ship. See the edit an existing image picture.
- •It fills space you asked it to leave. The left third of the banner is clear wall as requested, with a green 'Learn More' button painted into the photograph. Say 'no text or buttons' when you want room for your own. See the wide aspect ratio picture.
- •Headings exact, body invented. The four steps kept their wording and order, each with a line of instructions nobody wrote, and the layout alternates left and right rather than the plain list you might expect. See the infographic with steps picture.
- •Style prompts come signed. The ukiyo-e print carries a fabricated title cartouche, a place name in the corner and a red seal. Crop or repaint them before commercial use. See the specific art style picture.
- •Charts are right and slightly embellished. Six bars at the stated values in order, in a clean designed style, plus an axis label '(in thousands)' that the prompt never gave. See the chart picture.
- •Left-right placement was right and the pencil balanced on the rim, but the glass holds four ice cubes rather than three. Exact counts of small things remain the weak spot across the family. See the hard composition picture.
- •It is quick and even. Every one of the twelve pictures came back in 9 to 14 seconds at 1K, against 15 to 87 for Pro and about 3 for Lite.
- •1K sizes follow the shape: 1024 by 1024 at square, 896 by 1200 at 3:4, 1584 by 672 at 21:9. Ask for 2K or 4K when the long edge needs to pass about 1600 pixels.
Nano Banana 2 limits and API parameters
| Limit | Value |
|---|---|
| Output size | 512px, 1K, 2K or 4K in fifteen fixed shapes, including 4:1, 1:4, 8:1 and 1:8 strips. fal API reference, read |
| Reference images | Up to 10 objects, 4 characters and 3 styles in one request. |
| Transparent background | Not supported. fal API reference, read |
| Rate limits | Google meters images per minute per project tier and shows the live numbers in AI Studio rather than publishing a table. Gemini API rate limits, read |
›Every setting, for developers
fal (fal-ai/nano-banana-2)
fal API reference, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| prompt | string | Any text Required. | |
| aspect_ratio | enum | auto, 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16, 4:1, 1:4, 8:1, 1:8 | auto |
| resolution | enum | 0.5K, 1K, 2K, 4K | 1K |
| num_images | integer | 1 or more | 1 |
| seed | integer | Any integer Omit for a random seed. | |
| output_format | enum | jpeg, png, webp | png |
| safety_tolerance | enum | 1 to 6, 1 is strictest | 4 |
| system_prompt | string | Any text | empty |
| thinking_level | enum | minimal, high Unset means no thinking. fal adds $0.002 for high. | |
| enable_web_search | boolean | true, false fal adds $0.015 per request. | |
| image_urls | list of URLs | Reference images Edit endpoint only, where it is required. |
Gemini API (gemini-3.1-flash-image)
Gemini API image generation guide, read
| Parameter | Type | Values | Default |
|---|---|---|---|
| aspectRatio | enum | 1:1, 3:2, 2:3, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9 | |
| imageSize | enum | 512px, 1K, 2K, 4K | |
| reference images | inline images | Up to 10 objects, 4 characters and 3 style references | |
| thinkingLevel | enum | minimal, high | |
| grounding | tool | Google Search and Google Image Search |
What Nano Banana 2 will not draw
Google's Generative AI Prohibited Use Policy applies on every surface. It bars child sexual abuse material, non-consensual intimate imagery, content that facilitates violent extremism, self-harm or illegal activity, hate speech and harassment, and impersonating a real person without disclosure in order to deceive. There is no adult content mode. fal exposes a safety_tolerance dial from 1 to 6 but it operates inside the same policy. Every image is watermarked with SynthID.
Alternatives to Nano Banana 2
Frequently asked questions
What is Nano Banana 2?
Nano Banana 2 is Google's name for Gemini 3.1 Flash Image, released on 26 February 2026. It replaced the original Nano Banana as the default image model in the Gemini app and is the model free accounts generate with.
Is Nano Banana 2 free?
In the Gemini app, yes, with a daily limit Google does not publish. The Gemini API has no free tier for it; a 1K image costs $0.067 there and $0.08 on fal, read on 9 September 2026.
Nano Banana 2 or Nano Banana Pro?
Nano Banana 2 for almost everything: it is faster, half the price and close on quality. Nano Banana Pro when the brief has a lot of set text, many reference photos or needs 4K with the most detail.
What does the thinking setting do?
With thinking on high the model plans the image before drawing it, which costs a fraction of a cent and a few seconds. Our ladder above shows the same two prompts at minimal and high so you can judge whether it is worth it for your kind of prompt.
Does Nano Banana 2 have an API?
Yes. The model id is gemini-3.1-flash-image on the Gemini API and Vertex AI, and it is hosted on fal as fal-ai/nano-banana-2 and on Replicate.
Is Nano Banana 2 censored?
It runs under Google's Generative AI Prohibited Use Policy on every surface, so sexual content, real-person impersonation, violence and hate are blocked and there is no adult mode. fal's safety_tolerance dial from 1 to 6 loosens borderline filtering but not the policy. Every image carries a SynthID watermark.
How this page is made
Prices, limits and settings are read from the linked vendor and host pages on the dates shown. Every picture was generated by us through fal with the request shown under it, and the original files are kept unedited. A generation that failed is shown as a failure. Model and vendor names are used to identify the products; no vendor imagery appears on this page.
