Skip to main content

Nano Banana Pro API Quota Increase: Tier 1–3 Caps in Images

Nano Banana Pro API quota grows with your billing tier, not with extra keys: Tier 1's $10 per 10 minutes buys about 74 Pro images, Tier 3's $200 about 1,488.

LaoZhang AI TeamPublishedUpdated 12 min read
On this page
Nano Banana Pro API quota by billing tier: about 74, 372 and 1,488 Pro images per 10 minutes on Tiers 1 to 3

More Nano Banana Pro capacity on the Gemini API comes from the billing tier of the Google Cloud project that sends the requests. Link a billing account and you are on Tier 1. Pay $100 and wait 3 days after your first successful payment for Tier 2, or pay $1,000 and wait 30 days for Tier 3. Google upgrades the project automatically. If that still isn't enough, Google has a form for quota and billing-cap increases, and the Batch API gives you a separate, half-price queue for anything that can wait.

New API keys, a Google AI Pro or Ultra subscription, and the $300 Google Cloud Welcome credit add no API capacity. If you're hitting limits in the Gemini app or AI Mode rather than in your own code, that's a different quota system, covered in Nano Banana Pro Limits: Gemini, AI Mode, Workspace and API Quotas.

The model ID to call is gemini-3-pro-image. If your code still calls gemini-3-pro-image-preview, those errors have nothing to do with quota: Google shut that ID down on June 25, 2026.

Four limits that cap Nano Banana Pro API throughput

Google enforces four separate ceilings on image generation, and they fail in different ways. The fix depends on which one you hit, so it helps to tell them apart before you ask for more capacity.

LimitApplies toValuesWhen you hit it
Images per minute (IPM), requests per minute and per dayEach project, not each API keyLive values on the AI Studio rate limit page429 RESOURCE_EXHAUSTED; daily quotas reset at midnight Pacific
Spend per rolling 10 minutesYour usage tierTier 1 $10, Tier 2 $50, Tier 3 $200 (Gemini API rate limits page)429 RESOURCE_EXHAUSTED
Monthly spend capThe billing account, summed across every linked projectTier 1 $250, Tier 2 $2,000, Tier 3 $20,000–$100,000+ per month (Gemini API billing docs)Every linked project pauses until the 1st of the next month
Batch queueEach model, across all active batch jobsEnqueued-token table, shown in the Batch section belowNew batch jobs can't be queued

Google doesn't publish per-tier IPM or RPM numbers for Nano Banana Pro, including for Tier 3. The docs send you to AI Studio for your current values and add that "specified rate limits are not guaranteed and actual capacity may vary."

The spend-based limit applies depending on your billing history and account standing. When it does apply, it is usually the tighter ceiling for an expensive model like Nano Banana Pro, as the next section shows.

The monthly cap leaves a recognizable message. Developers on Google's forum quote it as "Your billing account has exceeded its monthly spending cap. Please go to AI Studio at https://ai.studio/billing to manage your billing". Two other stops look like quota problems but aren't. If anyone with editor or owner access set a project spend cap on the AI Studio Spend page, the project stops at that number. And on Prepay billing, a $0 credit balance makes every key in every project on that billing account fail with HTTP 402 until you add credits.

For the 429 itself, including how long to wait and how to retry, see Gemini Image 429 Rate Limit and Nano Banana Pro RESOURCE_EXHAUSTED.

How many Nano Banana Pro images each tier's caps allow

Spend caps are in dollars, so you can turn them into images by dividing by the per-image price. As of October 8, 2026, Google's pricing page lists these Standard prices:

  • Nano Banana Pro 1K or 2K: listed as $0.134. The exact figure is 1,120 output tokens at $120 per million, or $0.1344, and that is the price used below.
  • Nano Banana Pro 4K: $0.24.
  • Nano Banana 2.1 (gemini-nano-banana-2.1) 1K: $0.0336, which is 1,120 tokens at $30 per million.

The counts divide each cap by those prices and round down.

TierSpend per 10 minPro 1K/2K per 10 min (at $0.1344)Pro 4K per 10 min (at $0.24)Nano Banana 2.1 1K per 10 min (at $0.0336)Monthly capPro 1K/2K per month (at $0.1344)
Tier 1$107441297$2501,860
Tier 2$503722081,488$2,00014,880
Tier 3$2001,4888335,952$20,000 (minimum)148,809

For example, $10 ÷ $0.1344 = 74.4, so Tier 1 allows 74 Pro images per 10 minutes. Dividing by the rounded $0.134 list price gives the same 10-minute counts but slightly higher monthly ones: $250 ÷ $0.134 = 1,865 and $20,000 ÷ $0.134 = 149,253. These counts cover image output only. Your prompt, any reference images (560 input tokens each on Pro, about $0.0011) and thinking tokens add cost, so the real counts are slightly lower.

Two things stand out:

  • The 10-minute cap also limits your sustained rate. You can burst within a window, but averaged over time Tier 1 works out to about 7 Pro images a minute, Tier 2 about 37 and Tier 3 about 148, whatever IPM value AI Studio shows.
  • The monthly cap runs out in hours at full speed. $250 ÷ $10 = 25 windows, so a Tier 1 project generating at the 10-minute limit hits its monthly cap after about 4 hours 10 minutes. Tier 2 lasts 40 windows, about 6 hours 40 minutes. Tier 3's $20,000 floor lasts 100 windows, about 16 hours 40 minutes.

At 1K, Nano Banana 2.1 costs a quarter of a Pro image ($0.0336 × 4 = $0.1344), so it fits four times as many images into every cap: a Tier 1 month covers 7,440 of them ($250 ÷ $0.0336). Its thinking defaults to medium, and those tokens are billed on top at $7.50 per million, so treat that figure as a ceiling. Batch pricing halves the per-image cost, so Tier 1's $250 also covers 3,720 Pro 1K/2K images if they all go through Batch ($250 ÷ $0.0672). For full price tables, see Gemini 3 Pro Image API Pricing and Nano Banana 2.1 API Pricing.

Raising the tier: billing, $100 + 3 days, $1,000 + 30 days

Tiers belong to the billing account. Every project linked to that account inherits its tier, and every API key in a project inherits the project's tier and billing status. Moving a project to another billing account switches it to that account's tier.

  1. Get to Tier 1 by linking billing. In AI Studio, open Set up billing from the API keys page or the Projects page, choose or create a Cloud Billing account, and finish setup. New accounts default to Prepay with a $5 minimum purchase. Credits expire after 12 months. Free to Tier 1 usually happens instantly. A Free-tier project can't generate Nano Banana Pro images at all, because image output isn't available on the free tier for any Gemini image model.
  2. Reach Tier 2: $100 paid plus 3 days since your first successful payment.
  3. Reach Tier 3: $1,000 paid plus 30 days since your first successful payment.
  4. Check the result in the Billing Tier column of the AI Studio Projects page. Upgrades after Tier 1 take effect within about 10 minutes of qualifying.

Gemini API tier ladder from linked billing to Tier 2 and Tier 3, applied per project rather than per key

The spend that counts is total Google Cloud spend on the billing account, not only Gemini API usage. Google also notes that "in rare cases an upgrade request may be denied." There is no separate "T3 key" and no Tier 3 price for the model; the Gemini API Tier 3 (T3) guide walks through the upgrade and what to do if the tier doesn't update. gemini-3-pro-image costs the same on every paid tier. For Prepay, Postpay and the "No credits" status, see Gemini API Billing Tiers.

More images before the next tier: Batch, Nano Banana 2.1, a gateway

The 30-day wait for Tier 3 is fixed. While you wait, three options add real output, and each has a different trade-off.

Three ways to get more images before Tier 3: Batch at half price, Nano Banana 2.1 for drafts, and a paid gateway, plus why extra API keys add no capacity

Batch API: half price and a separate queue

Batch runs Nano Banana Pro at $0.067 per 1K or 2K image and $0.12 per 4K image. Batch requests have their own rate limits, separate from interactive calls: 100 concurrent batch requests, plus a cap on tokens queued per model. Batch spend still counts toward the monthly billing cap. As of October 8, 2026, the rate limits page lists these image-model queue sizes. It still uses the old preview names for two of them, even though those IDs were shut down on June 25, 2026. Google's pricing page also lists a Flex row for Pro at the Batch price, but the Flex documentation lists no image model, so plan on Batch as the documented half-price route.

Model as listed on the pageModel ID to callTier 1Tier 2Tier 3
Gemini 3 Pro Image Previewgemini-3-pro-image2,000,000270,000,0001,000,000,000
Gemini 3.1 Flash Image Previewgemini-3.1-flash-image (Nano Banana 2, deprecated)1,000,000250,000,000750,000,000
Gemini 3.1 Flash Lite Imagegemini-3.1-flash-lite-image2,000,000270,000,0001,000,000,000

Nano Banana 2.1 has no row on the page yet, so its Batch queue size isn't published. The queue is counted in tokens, including prompts and reference images, so the number of images it holds depends on your requests. The jump that matters is from Tier 1 to Tier 2: the Pro queue grows 135 times. For job setup and the 4K exception, see Nano Banana Pro Batch API Cost.

Nano Banana 2.1 for drafts

Not every image in a pipeline needs Pro. Drafts, variations and thumbnails can go to gemini-nano-banana-2.1, which became generally available on October 6, 2026. It costs $0.0336 per 1K image and $0.0504 per 2K image, so at 1K it fits four times as many images into the same spend caps. Save Pro for finals where the wording in the image has to be exact: in a 9-prompt test of Nano Banana 2.1, Pro and GPT Image 2.5, Pro kept the requested text, while 2.1 changed the wording in 3 of the text tasks (one image per prompt, so treat it as a signal rather than a rate).

If your drafts still run on Nano Banana 2 (gemini-3.1-flash-image), Google has deprecated it on the Gemini API and recommends 2.1 for new projects. No shutdown date has been announced. Switching cuts the 1K price from $0.067 to $0.0336. Nano Banana 2 Lite vs 2 vs 2.1 vs Pro covers which jobs suit which model, and Nano Banana 2 API Pricing has the old model's full price card.

A paid gateway as a separate contract

With a gateway, you pay the gateway per call instead of sending requests from your own Google project, so your project's tier no longer limits you. The gateway's own terms and limits apply instead. LaoZhang API lists gemini-3-pro-image at $0.09 per call for 1K, 2K or 4K, as of October 8, 2026. A response that returns HTTP 200 without an image still counts as one call, and size and aspect ratio can be set only on its Gemini-native route. It forwards requests to Google's model, so Google's safety filters still apply.

Disclosure: laozhang.ai is our own service. Its prices are platform pricing, not an independent comparison.

Asking Google for more: the Quota and Billing Cap Increase Request

The rate limits page links a paid tier rate limit increase request. It opens a Google Form called "Gemini API Quota and Billing Cap Increase Request". The same route covers monthly caps: the billing docs say that while default caps are preset, "you can request an increase." Google promises a review but no outcome: "We offer no guarantees about increasing your rate limit, but we'll do our best to review your request."

Use it when you consistently hit a limit during normal traffic on your current tier, or when Tier 3's $20,000 monthly floor is below what you'll spend. If you hit a spend-based 429 only once in a while, Google's own advice comes first: wait and retry, and reduce expensive requests.

When the tier doesn't update or drops back to Tier 1

The docs say upgrades are automatic. Developer threads on Google's AI Developers Forum show it doesn't always work that way:

  • Stuck at Tier 1. In an April 2026 thread, a team that had spent nearly $2,000 over more than a month still showed Tier 1 and then hit the $250 monthly cap. Another developer replied that submitting the Quota and Billing Cap Increase Request form had fixed the same problem for them more than once.
  • Dropped from Tier 3 to Tier 1. In August 2026, several billing accounts fell from Tier 3 to Tier 1 without notice, and production traffic started getting 429s. A Google staff member posted a separate downgrade report form on August 24 and said the next day that the reported accounts had been fixed. That form showed as closed on October 5, 2026.

If the Projects page shows a lower tier than your payment history supports, first check that you're looking at the billing account your project actually uses. Then file the increase request form with your project number. These are reports from developers, not documented behavior. With the downgrade form closed, the increase request form is the route still open.

What doesn't raise Nano Banana Pro API quota

  • More API keys in the same project. Limits apply per project, and every key's spend counts toward the same project and billing-account totals.
  • A Google AI Pro or Ultra subscription. It raises limits in the Gemini app and unlocks image models in the AI Studio playground. API keys take their tier from the project's Cloud Billing account, so the plan doesn't change your API quota. AI Studio's upgrade dialog lists "Gemini API: pay per request" and "Google AI: monthly subscription" as separate options.
  • The $300 Google Cloud Welcome credit. Google's billing FAQ says it can't be used for the Gemini API or AI Studio. Since March 2026, Gemini API costs are excluded from the Free Trial.
  • More projects to split one workload. A second project gets its own per-minute limits but shares the billing account's tier and monthly cap. Use separate projects for separate products, not to work around the limits on one product.

FAQ about Nano Banana Pro API limits

Is Nano Banana Pro free via the API? No. As of October 8, 2026, Google's pricing page lists image output as "Not available" on the free tier for every Gemini image model, including Nano Banana 2.1. You need Tier 1 billing before gemini-3-pro-image returns any images.

What is the Nano Banana Pro API quota limit per minute? Google doesn't publish a fixed number. Your project's live IPM and RPM appear on the AI Studio rate limit page. Averaged over time, the 10-minute spend cap also holds you to about 7 Pro 1K/2K images a minute on Tier 1, 37 on Tier 2 and 148 on Tier 3.

How can I get past the Gemini API rate limit for Nano Banana Pro? There's no legitimate bypass. You can move up a tier, send waiting work through Batch, put drafts on Nano Banana 2.1, file Google's increase request, or buy calls through a paid gateway under its terms. Extra keys in one project don't help.

Does Tier 3 make Nano Banana Pro cheaper? No. The per-image price is the same on every paid tier: $0.134 for 1K/2K and $0.24 for 4K, as of October 8, 2026. Only Batch halves it. A higher tier buys more images per 10 minutes and per month, not cheaper ones.

Sources6

External pages this guide links to, in the order they appear. Last updated Oct 8, 2026.

  1. 1.AI Studio rate limit pageaistudio.google.com/rate-limit
  2. 2.Gemini API rate limits pageai.google.dev/gemini-api/docs/rate-limits
  3. 3.Gemini API billing docsai.google.dev/gemini-api/docs/billing
  4. 4.ai.studioai.studio/billing
  5. 5.AI Studio Projects pageaistudio.google.com/projects
  6. 6.paid tier rate limit increase requestforms.gle/ETzX94k8jf7iSotH9
More in API Guides
Diagram connecting application JSON and YAML to an image API request, with a separate PDF document-processing step
API Guides

Nano Banana Pro API Guide: JSON, YAML, Image Saving, and PDFs

Call gemini-3-pro-image through Google Interactions, save image bytes from the full response, and compile your own JSON or YAML job settings. PDFs need a document-capable model before Pro receives the verified facts and page images.

12 min