Google’s Nano Banana 2.1 API is billed by tokens, not by a fixed per-image subscription. On the Standard paid tier, input text, images and video cost $1.50 per 1 million tokens, text or thinking output costs $7.50 per 1 million tokens, and generated image output costs $30 per 1 million image tokens. Using Google’s token estimates, an image-only request is approximately $0.0336 at 1K resolution, $0.0504 at 2K and $0.113 at 4K. Your actual bill can be higher when prompts contain images or video, when the model produces text, or when Search grounding is enabled.
The official model ID is gemini-nano-banana-2.1. It accepts text, images, video and PDFs, supports multi-turn editing and up to 14 reference images, and can be called through Google’s Interactions API or Batch API.
As an Amazon Associate I earn from qualifying purchases.
Nano Banana 2.1 pricing at a glance
Google lists Nano Banana 2.1 in the Gemini Developer API Standard paid tier. The following rates are the published 2026 prices and are denominated in US dollars.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallOutdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware match| Billable item | Standard paid price | What it covers |
|---|---|---|
| Input | $1.50 per 1 million tokens | Text, image and video tokens sent to the model |
| Output text and thinking | $7.50 per 1 million tokens | Text returned by the model and reasoning tokens, when produced |
| Output images | $30.00 per 1 million image tokens | Generated image data, measured with image-token units |
| Search grounding | 5,000 requests per month included; then $14 per 1,000 requests | Google Web and Image Search grounding shared across Gemini 3 and newer models |
The grounding allowance is shared rather than reserved exclusively for Nano Banana 2.1. One customer request can create one or more billable search queries, so the number of requests you submit is not always the number of search queries charged.
#1 Best Overall
- [Ultra-Lightweight Ear Buds Designed for Small Ears] Each earbud weighs only 3.7g and features a compact, ergonomic in-ear design made especially for small ears. Secure, low-profile, and comfortable for workouts, all-day wear or overnight sleeping.The ultra-slim profile sits flush in your ear with no hard edges poking or rubbing, so you can even sleep on your side without discomfort.
- [Immersive Stereo Sound with TOZO OrigX Technology] TOZO OrigX tuning delivers clear vocals, balanced mids, and natural stereo sound for music, podcasts, and videos.
- [Long Battery Life for Daily Use] Get up to 7 hours of playtime on a single charge, with up to 32 hours total using the charging case—ideal for workdays, commuting, and extended listening sessions.
- [Bluetooth 5.3 & Stable Connection] Bluetooth 5.3 provides fast pairing, stable wireless performance, and reduced dropouts as you move around home or office.
- [Deep Bass with Clear Vocals] High-performance drivers produce punchy bass while keeping vocals clean and detailed for everyday listening.
What one generated image actually costs
Google’s image-token estimates translate the $30-per-million output rate into approximate image-only prices:
| Output resolution | Approximate output image tokens | Estimated image-only cost |
|---|---|---|
| 1K | About 1,120 tokens | $0.0336 |
| 2K | About 1,680 tokens | $0.0504 |
| 4K | About 3,780 tokens | $0.113 (approximately $0.1134 before rounding) |
These are estimates for the generated image output alone. A request with a text prompt also consumes input tokens. Every input image consumes about 1,120 input tokens, and video consumes tokens according to its content. If the response includes explanatory text or thinking, those output tokens use the separate $7.50 rate. Search grounding adds its own charge after the shared monthly allowance is exhausted.
Why the 1K estimate is $0.0336
The calculation is straightforward: 1,120 output image tokens divided by 1,000,000, multiplied by $30, equals $0.0336. The same method gives $0.0504 for 1,680 tokens at 2K and about $0.1134 for 3,780 tokens at 4K. Google publishes these as approximate equivalents, so treat them as budgeting figures rather than a guaranteed invoice for every request.
Free tools Windows power users keep installed
One-click scans. No signup required.
Examples of input-image costs
One reference image uses about 1,120 input tokens. At $1.50 per million input tokens, that is approximately $0.00168 for the image input itself. Fourteen such images would account for about 15,680 input tokens, or roughly $0.02352, before counting the text prompt and generated output. The resulting image would still add approximately $0.0336, $0.0504 or $0.113 depending on resolution.
Is Nano Banana 2.1 free?
Google’s general Gemini API page uses “start building free of charge” language, but the Nano Banana 2.1 model row shows that a Free Tier is not available. Do not plan on unlimited no-cost API generation for this model. Production use requires the Standard paid billing path, with charges based on the tokens and services used.
Rank #2
- 【IP68 Waterproof Multifunctinal Smart Watch】Torjalph fitness activity tracker smart watch is IP68 waterproof technology ,you can wear it when swimming,handwashing,it's rain-proof which means that you perform multiple exercise modes outside without worrying about sweat or rain corroding your fitness smart watch,and this watch equipped with a 230 mAh large capacity battery magnetic charging ,and this sports activity tracker watch can be used 5-7 days If fully charged.
- 【Larger 1.83'' Full Screen Watch & Multi-dial Selection】1.83 inch full HD color touch screen smart watch for men / women provides excellent 240*284 resolution image quality ,This Bluetooth smart watch offers a better touch experience and remarkable immersion and ,more than 200 watch faces for you to choose from App"Da Fit ", you can change the dial style according to your mood to wear keeping the world on your wrist
- 【Accurate Activity Tracker Smart Watch 】sports watch activity tracker uses sweat and dust resistant breathable silicone strap, our Fitness Tracker Smart Watch for Android iOS Phones supports 12 sports modes,walking,running,bicycle,swimming,badminton,basketball,football,rope skipping. This bluetooth calls smart watch for men women is lightweight and stylish, bringing more convenient and relaxing sports experience while also providing more protection.
- 【Your 24 Hour Personal Daily Health Monitor】Fitness watch for women men full touch screen IP67 waterproof bluetooth smartwatch for Android iOS Phones with heart rate blood pressure monitor multifunction Andriod smart watch with steps counter sleep monitor sports fitness watch, monitoring blood oxygen saturation allows you to understand your physical condition in real time to track fitness progress ,to know your health status in time.
- 【More Functions Bluetooth Call Smart Watch】This fitness activity smart watch full screen for iPhone and Android phones includes a bunch of new features including:The bluetooth smart watch can hang up the phone, receive and display WhatsAPP /Text /Facebook /Twitter /Messenger reminder.Support sedentary reminder, alarm clock, stopwatch, timer,weather information,music control, remote camera control compatible with Android 5.0 and above or iOS 9.0 and above smartphones.
This does not establish a consumer-app subscription allowance. Google’s published model and API pages do not provide a current Nano Banana 2.1 quota table for a consumer application, so an app subscription should not be treated as equivalent to API access.
Model limits and supported inputs
The stable model identifier is gemini-nano-banana-2.1. Google lists the following context and output limits.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Fix the driver behind crashes, sound loss and screen glitches3Clear out junk files and repair common Windows errors| Capability or limit | Nano Banana 2.1 detail |
|---|---|
| Input modalities | Text, images, video and PDFs |
| Output modalities | Images and text |
| Maximum input context | 131,072 tokens |
| Maximum output context | 32,768 tokens |
| Maximum uploaded input size | 500 MB |
| Reference images | Up to 14 images per prompt |
| Image resolutions | 1K, 2K and 4K |
| Aspect ratios | From square 1:1 through panoramic ratios listed by Google |
| Batch processing | Supported, with turnaround of up to 24 hours |
Reference-image behavior
The 14-image ceiling is a total per prompt, not 14 images of every type. For Nano Banana 2.1, Google describes up to four reference images for preserving characters and up to ten for preserving objects within that overall limit. For example, a character-consistency request might use four identity references and reserve the remaining slots for scene or prop references, but the combined count cannot exceed 14.
Features that are not available
The model supports image generation, Search grounding and thinking. It does not support audio generation, function calling, structured outputs, the Live API or URL context. Parameters such as seed, topK, logprobs, temperature and topP are explicitly unsupported; sending them returns an API error. Remove those fields instead of retrying the same request unchanged.
Calling Nano Banana 2.1 from Python
Google’s image-generation guide uses the Interactions API. Set the model to the exact ID, provide a text prompt, then decode the returned base64 image data and write it as a PNG.
Rank #3
- [Low Noise High Fidelity] New bee guitar cord with 22AWG oxygen-free-copper center conductors provides enhanced signal transfer, the PVC insulation material and the high-density stranded copper double shield provide noise rejection for quiet operation
- [Ideal for Choice] New bee 1/4 straight and right angle metal connectors, ensures a secure connection in any instrument, pedal, or amp jack. L shape great fro on stage guitar stands
- [Wide Compatible] Suitable for mixer, power amplifier, microphone, effector, decoder, equalizer, electoronic organ, electric guitar, electric bass and other electronic equipments. Widely used in the installation of smart home, karaoke speaker, professional stage audio, etc
- [Easy Disassemble] Making incidental repairs quick and easy! Very nice, flexible cable
- [Dont Hesite to Order] 24 hours customer services and professional technology team are standing by
import base64
from google import genai
client = genai.Client()
interaction = client.interactions.create(
model="gemini-nano-banana-2.1",
input="Create a clean product illustration of a red hiking backpack on a white background.",
)
for item in interaction.output:
if getattr(item, "type", None) == "image":
image_bytes = base64.b64decode(item.data)
with open("backpack.png", "wb") as file:
file.write(image_bytes)The response can contain more than one output item. Check the item type before decoding so that a text or thinking item is not mistaken for image data. The exact SDK object shape can change between library releases, so keep your code aligned with the current Google SDK documentation.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Calling Nano Banana 2.1 from JavaScript
In JavaScript, initialize the Google GenAI client and create an interaction with the same model ID.
import { GoogleGenAI } from "@google/genai";
const ai = new GoogleGenAI({});
const interaction = await ai.interactions.create({
model: "gemini-nano-banana-2.1",
input: "Create a clean product illustration of a red hiking backpack on a white background.",
});
for (const item of interaction.output ?? []) {
if (item.type === "image") {
const imageBuffer = Buffer.from(item.data, "base64");
// Save imageBuffer as a PNG with your preferred file system or storage API.
}
}Keep API credentials on a server or in a protected runtime. Do not embed a production key in browser JavaScript, a public repository or a downloadable application.
Using reference images, video and PDFs
Text-only prompts are the simplest and least expensive requests to estimate. When you add reference images, include their input-token cost in the budget. The model can also accept video and PDF inputs, but their tokens count as input usage and can consume the 131,072-token context window. A 500 MB upload limit applies to the input payload, yet a file below 500 MB can still exceed the token limit after it is processed.
For iterative work, use multi-turn editing rather than resending an unnecessarily large prompt every time. Each turn is still a billable request, and any images or video included again consume input tokens. Keep only the references needed to preserve identity, object details or composition.
Rank #4
- Digital Family Calendar: BSIMB digital calendar wall touch screen puts an end to your chaotic life. Designed for busy families, the digital calendar keeps you organised and in step with your family. Assign different colours and tasks to each family member to make family scheduling easier and more intuitive.
- Easy Setup and Auto-Sync: Plug in, connect to WLAN and auto-syn your electric calendar in minutes. Digital wall calendar compatible with Google calendar, iCloud calendar, Outlook calendar, Cozi calendar and Yahoo calendar. Add events directly on the device or through the ecalendar app. Electronic calendar stays current whether you're at work or traveling—no manual updates needed.
- All-In-One Home Assistant: Smart calendar builds habits and encourage independence with personalised chores. Display daily meal plan to reduce the daily ‘What's for dinner?’ stress. To-do lists keep track of everything you want. Turning chores into valuable achievements with stars and rewards, making life easier and more enjoyable.
- Manage Anytime Anywhere: BSIMB electronic calendar for family stay connected with the powerful ecalendar app. Access and manage your digital wall calendar anytime, anywhere with free features like stars & rewards, photo screensaver, chore charts, meals plans, to-do lists etc. Electronic planner paid features like Magic Import tailor advanced tools for users with special needs.
- Stay in Touch with Loved Ones: Share photos or videos from your phone to electronic wall calendar with app, instantly update your life moment, and keep in touch with your family, especially those who are far away; you can also surprise your smart calendar digital wall by sending him a wish card; it's also great for keeping tech-illiterate elders in touch with the youngsters.
Standard API versus Batch API
The Standard path is intended for interactive responses. For large sets of independent image jobs, Google supports the same image-generation capabilities through its Batch API. Batch trades speed for capacity: Google documents higher rate limits with turnaround of up to 24 hours.
| Path | Best fit | Trade-off |
|---|---|---|
| Standard | Interactive editing, previews and user-facing tools | Subject to normal project quotas and minute-based limits |
| Batch | Backfills, catalogs and other bulk jobs that can wait | Results may take up to 24 hours, despite higher limits |
Rate limits, quotas and 429 errors
Quotas are measured at the Google Cloud project level, not separately for each API key. Creating several keys in one project does not multiply the project’s capacity. Google documents limits in several dimensions:
- RPM: requests per minute.
- Input TPM: input tokens per minute.
- RPD: requests per day.
- IPM: images per minute for image-capable models.
Daily request quotas reset at midnight Pacific time. Exceeding any applicable limit returns a rate-limit error. A 429 response with RESOURCE_EXHAUSTED can also mean that your spend rate, rather than your request count, is too high.
Spend-based rolling limits
Google also documents rolling 10-minute spend limits of $10 for Tier 1, $50 for Tier 2 and $200 for Tier 3. These are rate protections, not monthly budgets. If a burst of expensive 4K generations triggers a 429, wait briefly, reduce the rate of costly requests and retry with backoff. If your workload consistently needs more capacity, request a quota increase when your project is eligible.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Practical retry strategy
- Classify the error as quota, spend-rate or invalid-parameter related. Do not retry an unsupported parameter error unchanged.
- For a temporary 429, pause briefly and retry with exponential backoff and jitter.
- Limit concurrency so that simultaneous jobs do not exhaust IPM or spend-rate limits.
- Queue non-urgent jobs for Batch API processing.
- Track input tokens, output resolution, grounding queries and HTTP status codes per project.
How to budget a production workload
Start with the number of images, their resolution and the amount of input material. For example, 1,000 image-only 1K generations are approximately 1,000 × $0.0336, or $33.60. The same count at 2K is about $50.40, while 4K is about $113. Add prompt and reference-image input charges, any text or thinking output and grounding charges.
A more realistic reference-heavy job might include four input images, a text prompt and one 2K output. The output estimate is $0.0504. The four references account for about 4,480 input tokens, costing approximately $0.00672 at the input rate, before the prompt text itself. This illustrates why resolution usually dominates image-only cost, while very large reference sets and video can materially increase input usage.
Controls that prevent surprise bills
- Choose 1K for drafts and reserve 2K or 4K for approved assets.
- Set application-level per-user and per-project budgets in addition to Google Cloud billing alerts.
- Record requested resolution and reference count with every job.
- Use Batch for work that does not need an immediate response.
- Disable Search grounding unless current web information is genuinely needed.
- Reject oversized files and unsupported generation parameters before sending a request.
What Nano Banana 2.1 is best suited to
The model’s combination of conversational editing, multiple reference images and 1K-to-4K output makes it suitable for workflows that need visual continuity: product variations, character-preserving scenes, object-focused edits and iterative art direction. The 14-image limit and separate character/object guidance are useful when consistency matters, while the 131,072-token context allows substantial text and file context.
It is less suitable when an application requires audio output, function calling, structured JSON responses, Live API sessions or URL-context retrieval, because those features are not supported by this model. In such systems, keep image generation as a dedicated step and handle orchestration, validation and metadata in your own application.
Recommended Free Tools
Bottom line on Nano Banana 2.1 cost
Budget approximately 3.36 cents for a 1K image, 5.04 cents for a 2K image and 11.3 cents for a 4K image when considering output alone. The API is not listed as free-tier eligible, and the final charge rises with input images, video, PDFs, text or thinking output and Search grounding. Use the exact model ID, stay within the 14-reference-image and 131,072-input-token limits, design around project-level quotas and move bulk work to Batch when a 24-hour turnaround is acceptable.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
