Applied artificial intelligence, neural systems, and smart agents.
Fusing generative AI, multi-agent frameworks, and high-throughput data processing to automate business workflows.
Nano Banana Pro
Nano Banana Pro is the gemini-3-pro-image unit: a 4K resolution ceiling, six-image object reference capacity and five-image character reference capacity, at a $0.134 rate per 1K or 2K image. Its core constraint is that Nano Banana 2, Google own lower-cost and newer unit, ranks level or above it on both Arena benchmarks at half the rate.
- up to 4K
- six object and five character references
- $0.134 per 1K image
Last checked: This is a reference page. It is re-checked against the vendor sources and updated when a new version ships. Change log
Google most expensive configuration, logging fewer votes
On the text-to-image benchmark Nano Banana Pro scores 1246 and Nano Banana 2 scores 1264. On the editing benchmark it is 1389 against 1385, which is a statistical tie in practice. Set the pricing specification beside that: Pro runs $0.134 for a 1K image and Nano Banana 2 runs $0.067 for the same size. Exactly double, for output not preferred on average.
The two benchmarks must be read separately, since one gap is real and the other is not. On text-to-image the distance is about eighteen points and the confidence intervals do not meet: Nano Banana 2 begins at 1258.7 and Pro ends at 1249.4. Pro is genuinely behind on that axis. On editing, Pro is four points ahead and the intervals overlap in full, which indicates a tie rather than a win. Summed up: Pro loses one benchmark, draws the other, and costs twice as much.
So under what conditions does Pro justify deployment
In two places, both documented in Google own specification. First, 4K: the four-kilo tier runs $0.24 on Pro and $0.151 on Nano Banana 2, so the higher the resolution tier, the smaller the price gap, and at the top tier the ratio is no longer double. Second, character references: Pro accepts up to five character images and Nano Banana 2 accepts four. For a workload of one scene with several people who must stay consistent, that one extra reference image may be the only requirement.
A reference count that runs counter to expectation
Google publishes two separate figures per unit: object images and character images. Pro provisions six objects and five characters, Nano Banana 2 provisions ten objects and four characters, and Nano Banana 2 Lite provisions fourteen objects. So the lowest-cost unit accepts the most objects and the highest-cost unit accepts the most characters. Our table builds the reference column from the object count so the two Google units can be compared at all, and that is precisely what places Pro low in that column. Its character figure is recorded here because it does not fit the table structure.
How many pixels does 4K represent
Google does not publish this figure. The image-generation specification states "up to 4K" for Pro and nowhere prints pixel dimensions for the 1K, 2K or 4K settings. For contrast: OpenAI publishes a 3840-pixel ceiling on the long edge and a total pixel count between 655,360 and 8,294,400 for its own unit. That is why the category table carries no resolution column at all: a comparable column cannot be built from one label and one figure.
From Iran
The Gemini API available-regions document was read today and Iran is not recorded on it. This is the same document checked for every other Google API unit, and it makes no statement about the Gemini application.
Specifications
Every number here comes from the vendor's own page, with that page linked beside it.
Using it from Iran
This column carries a date and says how it was checked, because most listicles guess it. Where we have only read a vendor policy page, the note below says exactly that.
- Reachable
- blocked
- Payment
- no working route
- Free tier
- no
- Measured on
How we checked: The Gemini API available-regions page lists the countries in which the interface is active, and Iran is not recorded on that list. This finding comes from reading a document, not from a network measurement.
Documented strengths
- A 4K resolution ceiling, and at that tier the price gap with the lower-cost unit narrows
- The highest character reference capacity of any Google image unit: five images
- Google itself describes its text rendering as advanced, for infographics and menus
- Deploys on the same API surface as the rest of Gemini, with no separate integration to learn
Room for improvement and technical considerations
- Nano Banana 2 scores level or higher on both Arena benchmarks at half the rate.
- Google publishes no pixel dimensions for 4K, so 4K here functions as a label rather than a measurable value.
- It trails its own lower-cost unit on object reference capacity: six against ten.
- No weights are published, and no local deployment route exists.
- Iran is not recorded on the Gemini API available-regions list.
- It carries two rows with two distinct scores on both benchmarks, meaning the score follows the serving configuration, not the unit alone.
Technical verdict
If operating cost is not the constraint and a workload requires several consistent characters in one scene, provisioning Pro is justified. In every other case, Nano Banana 2 delivers the same output at half the cost, and the recorded evidence supports this. This is not a preference statement: if the leaderboard shifts, this assessment updates with it.
Frequently raised questions, with documented answers
What is Nano Banana Pro, technically
It is the commercial name for the Google image unit gemini-3-pro-image. Its resolution ceiling is 4K, its reference capacity is six object images and five character images, and its deployment surface is the ordinary Gemini API.
What is the pricing structure of Nano Banana Pro
$0.134 per 1K or 2K image and $0.24 per 4K image, on the standard tier. The batch-processing route runs at half of those rates.
What is the architectural difference between Nano Banana Pro and Nano Banana 2
Pro maps to gemini-3-pro-image and 2 maps to gemini-3.1-flash-image, meaning 2 is the newer unit. Pro provisions more character references and costs twice as much; 2 scores level or higher on both Arena benchmarks.
Is an access route available from Iran
Iran is not recorded on the Gemini API available-regions list, so no official route exists. This platform does not sell accounts and does not propose a route around this restriction.