Skip to content

Business application

DualDrive AI

Every AI model. Two routes. One app.

DualDrive AI is a Windows app that sends your tasks to the best AI models in two ways: through the API — in parallel, fast and with the exact cost — or through a chat window in your Chrome browser, with no API key and no token charges. One app instead of several subscriptions and five separate panels.

from $10 net/month

from $10 gross/month

  • Windows 10 (from 1903) and 11; browser mode: Google Chrome
  • License for 1 computer · key and installer link by email
  • Requires a constant internet connection (license verification)
The Task tab in DualDrive AI: six product-description prompts separated by #, the Parallel 6 setting, the GPT-6 Sol model and a batch cost estimate

Key capabilities

  • Up to 16 at once

    A batch of tasks in parallel

    The whole batch sent with one click, results appear as they arrive.

  • To the token

    Pay for what you use

    With the API you pay the provider only for what you send — no several subscriptions at once.

  • PLN 0 for tokens

    Browser mode

    A batch in the chat window on your own account in Chrome, with no API key.

  • 5 providers

    All models in one window

    ChatGPT, Claude, Gemini, Perplexity and DeepSeek — the best model for every task.

  • 1 → N

    Improve prompt

    A short idea turned into polished prompts, ready to send as a batch.

  • On your PC

    Your data stays with you

    History, answers and files stored locally; API keys encrypted by Windows.

Speed

A batch of tasks run in parallel

Instead of sending prompts one after another, you paste the whole batch — DualDrive AI sends them all at once. Results appear as they arrive, and when a provider asks you to slow down, the app slows down by itself and retries the task — nothing gets lost.

  • 4–8 tasks at once by default, depending on the provider, up to a maximum of 16
  • Automatic retry after a rate-limit refusal (429)
  • 6 product descriptions with the GPT-6 Sol model — the whole batch PLN 0.048
A batch running in DualDrive AI: the progress bar showing Done 2 of 6, in progress 4, with the first finished answer beside it

Improve prompt

One sentence → three polished prompts

With one click you turn a short idea into a precise prompt: role, context, requirements and output format. A request for “3 different images” or “10 descriptions” is split by the app into separate, self-contained tasks — a ready batch to send.

  • Automatic split into a batch when you ask for N separate outputs
  • Optional clarifying questions
  • Prompt history with search
The Improve prompt tab in DualDrive AI: the instruction Generate 3 different images of cats turned into three detailed prompts separated by #

Two routes

API or chat window? You choose each time you send

The API is faster — many tasks at once, paid per token. The chat window in your browser works on the account you already have — no API key and no token costs, but tasks run one after another. You choose each time you send, or set a default route.

  • API: ChatGPT (OpenAI), Claude (Anthropic), Gemini (Google), Perplexity and DeepSeek
  • Each provider with its own key, models, price list, history and balance
  • A default setting: “use automatically”
The How to ask the model? window in DualDrive AI with two options, Via the API and In browser mode, with a description of the differences

Browser mode

A batch in the chat window, with no token costs

The Chrome extension types the questions into the chat window on your account, waits for the answers and saves the results — text and PNG images — to disk. Example from the screenshot: 3 cat images from an improved prompt, generated in the ChatGPT window — PLN 0 in tokens; the same batch via the API is about PLN 2.07 by the app's estimate.

  • ChatGPT, Claude, Gemini, Perplexity and DeepSeek in the chat window
  • Choice of the chat window's model and thinking level
  • Within the plan and limits you already have with each service
The Task tab in DualDrive AI in image mode: the result of a batch of three cat images generated in the ChatGPT chat window, cost 0

Cost control

Know what you spend — before and after

Every call with its date, model, tokens, cost and route — API or browser. Search, filters, CSV export and a batch breakdown, with a chart of costs day by day split by model alongside.

  • An estimate before sending, including “1,000 such requests”
  • Account balance, monthly forecast and a warning threshold
  • Amounts in PLN, EUR or USD (ECB rates), net or gross
The History tab in DualDrive AI: a table of requests with date, model, method (API or Browser), tokens, cost and time, with a batch summary below

Files

Every answer saved to disk

Answers go automatically to the working folder in the right format — text, document, spreadsheet, image or vector graphic. The Files tab shows a preview and image thumbnails, so you never have to copy anything by hand.

  • Formats txt, md, docx, xlsx, csv, png, svg and more
  • SVG vector graphics with export to PNG
  • Your own working folder
The Files tab in DualDrive AI: the working folder, a list of saved answers and a preview of the selected file

Pricing

Choose a plan: a one-off license or a subscription with updates.

Recommended

Monthly subscription

Billed monthly

$10 net/month

$10 gross/month (incl. 23% Polish VAT)

Prepaid monthly access. Before the period ends you receive a payment link – no automatic charges.

  • Prepaid monthly, no automatic charges
  • Updates and support included
  • Cancel renewals at any time

Annual subscription

Billed annually

$40 net/year

$50 gross/year (incl. 23% Polish VAT)

Prepaid annual access. Before the period ends you receive a payment link – no automatic charges.

  • Prepaid once a year, no automatic charges
  • Updates and support included
  • Cancel renewals at any time

One-off license

One-off payment

$70 net

$80 gross (incl. 23% Polish VAT)

A single payment for the license, with no recurring charges.

  • License with no billing periods
  • One-off purchase, no commitment

Renewal via a payment link before the period ends – no automatic charges; you can cancel at any time.

Gross prices include Polish VAT. For businesses and customers from other countries VAT is calculated in the order summary (e.g. reverse charge in the EU, 0% outside the EU).

The price shown here is an estimate: we convert from PLN using the National Bank of Poland's exchange rate and round up. Your invoice and any payment you make are always processed in PLN - your card issuer or bank handles the actual conversion, and may apply its own rate or fees.

A one-off license is granted for an indefinite period: the application works for as long as we run the service and the license server. You are buying a license to use the application, not ownership of it. Details: Terms of sale, § 9.

How buying works

  1. Choose a plan and click “Buy now”.
  2. Enter your invoice details and pay online.
  3. You'll get your license key and the installer link by email.
  4. Install the app, enter the key, and get to work.
  • Online payment
  • VAT invoice
  • Access once paid
  • Support on business days

Screenshots

Real screenshots from the app. Click to enlarge.

  • The Task tab in DualDrive AI: six product-description prompts separated by #, the Parallel 6 setting, the GPT-6 Sol model and a batch cost estimate

    A batch of tasks

    Six tasks, one click — with the price before you send.

  • A batch running in DualDrive AI: the progress bar showing Done 2 of 6, in progress 4, with the first finished answer beside it

    Batch in progress

    Tasks run in parallel — results arrive as they come.

  • A batch result in DualDrive AI: a Sencha tea description, the Result 1 of 6 switch, the cost of a single request and of the whole batch

    Batch results

    Each result separately, with the exact cost.

  • The Improve prompt tab in DualDrive AI: the instruction Generate 3 different images of cats turned into three detailed prompts separated by #

    Improve prompt

    One sentence → three polished prompts.

  • The How to ask the model? window in DualDrive AI with two options, Via the API and In browser mode, with a description of the differences

    API or browser

    API or chat window? You choose each time you send.

  • The Model for browser work window in DualDrive AI: the Claude service selected, the Claude Opus 5.5 model and a high thinking level

    Model in the browser

    Claude Opus, GPT-6, Gemini 4 — on the account you already have.

  • The Task tab in DualDrive AI in image mode: the result of a batch of three cat images generated in the ChatGPT chat window, cost 0

    A batch of images

    A batch of images in the ChatGPT window — no token charges.

  • The History tab in DualDrive AI: a table of requests with date, model, method (API or Browser), tokens, cost and time, with a batch summary below

    History

    Full history: model, route, tokens and cost.

  • The Chart tab in DualDrive AI: a bar chart of the daily cost over 30 days and a table breaking the cost down by model

    Cost chart

    Know what you spend — day by day and model by model.

  • The Files tab in DualDrive AI: the working folder, a list of saved answers and a preview of the selected file

    Files

    Every answer saved to disk, in the right format.

  • The AI Settings tab in DualDrive AI: provider selection, account balance, currency and VAT, API keys and the model price list

    AI settings

    Five AI providers in one app.

  • The AI Browser tab in DualDrive AI: choice of the default chat-window model and instructions for loading the Chrome extension

    AI Browser

    Browser mode: default model and extension in three steps.

Use cases

  • Product descriptions for an online shop — dozens of descriptions in one batch, with the cost known before sending
  • A batch of images — pictures from improved prompts, via the API or in the ChatGPT chat window with no token costs
  • Marketing copy, posts and articles — the same brief in several variants, each saved as a separate file
  • Search with answers that cite their sources — tasks for Perplexity and models with web search
  • Working with documents — attached PDFs, Word, Excel and PowerPoint files as material for the task
  • Cheap bulk tasks — long batches with the DeepSeek model or in browser mode

Technical specifications

System
Windows 10 (from version 1903) and Windows 11
Browser
Google Chrome — only for browser mode; you load the extension once, and the app guides you through 3 steps
Internet
Required at all times — without a connection to the license server, features lock after about 2–3 minutes and come back on their own
Installation
A single installer for one user, with no administrator rights and no extra programs to download
AI providers
ChatGPT (OpenAI), Claude (Anthropic), Gemini (Google), Perplexity, DeepSeek — via the API with your own key or in the chat window on your account
AI costs
API charges are collected by the AI provider according to its price list — they are not included in the price of the app; browser mode has no token costs, within your plan
Interface language
Polish, English, German, French, Spanish, Italian
License
A key with online activation, one computer; transfer from your customer account once every 30 days
Privacy
History, answers and files stored locally; requests go only to your chosen AI provider, and license data and ECB exchange rates are also sent over the internet

Questions about DualDrive AI

Do I need an API key?

No. In browser mode, tasks go through the chat window on your account — all you need is Chrome and to be logged in to the service. An API key comes in handy when you want to work in parallel and as fast as possible.

How much does working through the API cost?

As much as you use — at the rates of the AI provider you pay directly. The app shows an estimate before sending and the exact cost after every task. For example, 6 product descriptions with the GPT-6 Sol model cost PLN 0.048.

Does the price of the app include API charges?

No. The price of DualDrive AI is a license for the app. You settle API charges directly with your chosen AI provider (OpenAI, Anthropic, Google, Perplexity, DeepSeek) according to its price list; browser mode does not use API tokens.

How does browser mode differ from the API?

The API is faster (many tasks at once) and paid per token. Browser mode has no token charges but is slower — tasks run one after another, in the chat window on your account, within the limits and terms of your plan with that service.

Which models are available?

All those the provider offers through the API for an ordinary key — including GPT-6 Astra, Sol and Luna, GPT Image 2.5, Claude Fable 5.1, Opus 5.5, Sonnet 5.5, Haiku 5.5, Gemini 4 Argon, Nano Banana 2.1, DeepSeek V4 and the Perplexity tiers. The list updates along with the price list.

Can I generate images?

Yes — through the API (GPT Image, Gemini Nano Banana) or in the ChatGPT chat window in browser mode. Every image is saved as a PNG file.

What system does it run on?

Windows 10 (from version 1903) and Windows 11. A single installer, no extra programs. Browser mode requires Google Chrome and loading the extension once — 3 steps, and the app guides you through them.

Does the app work without internet?

No. Working with AI models needs a connection, and the license is checked online — without a connection to the license server, features lock after about 2–3 minutes and come back on their own once the connection is restored.

Where does my data go?

To your disk. History, answers and files stay on your computer, and API keys are encrypted with the Windows mechanism (DPAPI). Requests go only to your chosen AI provider — just as when you use its website.

How do I move the license to a new computer?

The license works on one computer. You can transfer it from your customer account on the website — no more than once every 30 days.

Which purchase options are available?

A monthly or annual subscription, prepaid — before the period ends you receive a payment link, with no automatic charges — or a one-off license, for an indefinite period while the service is running.

Full feature list

Tasks and batches

  • A batch in one field – tasks separated by the # character, sent with one click.
  • Parallel execution – up to 16 tasks at once, a safe default pace for each provider, automatic retry after a rate-limit refusal (429).
  • Four kinds of task – text, image, SVG vector graphics (with export to PNG) and web search.
  • Attachments – images, PDFs, Word, Excel and PowerPoint documents, text files and code.
  • Model effort level – thinking level chosen per task (OpenAI, Claude, Gemini, DeepSeek); batch results browsed with arrows and copied with one click.

Improving prompts

  • Improve prompt – a short idea turned into a precise prompt: role, context, requirements, output format.
  • Splitting into a batch – a request for N separate outputs (e.g. 3 images, 10 descriptions) gives N ready, self-contained prompts.
  • Questions and history – optional clarifying questions and a prompt history with search.

Two routes to the models

  • API – ChatGPT (OpenAI), Claude (Anthropic), Gemini (Google), Perplexity and DeepSeek — each with its own key, models, price list, history and balance.
  • Browser mode – the same services through the chat window in Chrome, on your account; choice of the chat window's model and thinking level.
  • Choosing the route – each time you send, or a default setting: “use automatically”.

Costs and billing

  • Cost of every call – from the tokens actually used; an estimate before sending, including for “1,000 such requests”.
  • History and chart – search, filters, CSV export; costs and tokens day by day, split by model.
  • Balance and currencies – account balance, monthly forecast, warning threshold; PLN, EUR, USD (ECB rates), net or gross.
  • Model price list – updated with one click or automatically; models withdrawn by the provider disappear from the list on their own.

Files and formats

  • Automatic saving – every answer in the right format: txt, md, docx, xlsx, csv, png, svg and more.
  • Files tab – preview and image thumbnails; your own working folder.

Security and the program

  • Encrypted API keys – by the Windows mechanism (DPAPI) — once saved, only the last 4 characters are visible.
  • Local data – history, answers and files on your computer; the extension communicates only with the app on your computer.
  • Interface – 6 languages (Polish, English, German, French, Spanish, Italian), scaling and your own color schemes.

About the application

Anyone who works with AI every day usually pays for several subscriptions at once — ChatGPT for texts and images, Claude for writing and code, Gemini, Perplexity for search. On top of that they copy prompts one by one, wait for every answer and save the results to files by hand. And when 50 product descriptions or 100 images are needed, hours of clicking begin. DualDrive AI turns all of that into a single run.

How it works

  1. Describe. Type a task or a whole batch — separate each task with the # character.
  2. Improve. With one click, turn a short idea into a polished prompt; a request for “3 different images” is split by the app into 3 separate tasks.
  3. Choose your route. The API (in parallel, pay per token) or the chat window in your browser (no key and no token costs).
  4. Collect the results. Every answer is saved to disk in the right format, with cost, tokens and time recorded in the history.

Speed: a batch of tasks run in parallel

Instead of sending prompts one after another, you paste the whole batch — DualDrive AI sends them all at once (4–8 at a time by default, depending on the provider, up to a maximum of 16). Results appear as they arrive, and when a provider asks you to slow down, the app slows down by itself and retries the task — nothing gets lost. Example from the screenshot: 6 product descriptions with the GPT-6 Sol model — the whole batch cost PLN 0.048.

The best models without a subscription

With the API you pay the provider only for what you actually use — with no monthly fee for each service separately. In one app you get models from five providers and pick the best one for the task: GPT Image and Nano Banana for images, Claude for texts and code, Perplexity for answers with sources, DeepSeek for cheap bulk tasks. According to the price list built into the app, 1,000 product descriptions with the GPT-6 Sol model cost about PLN 7.77 — the app calculates this before sending.

Browser mode

Already have an account with ChatGPT, Claude, Gemini, Perplexity or DeepSeek? DualDrive AI can send a batch of tasks through the chat window in your Chrome browser — the extension types in the questions one after another, waits for the answers and saves the results (text, PNG images) to disk. No API key and no token charges — within the plan and limits you already have. It is the route for long batches, such as hundreds of images or descriptions, when cost matters more than seconds.

Full cost control

An estimate of tokens, cost and time before sending, the actual cost after every task, a history split into API and browser, a daily chart, an account balance with a low-balance warning, amounts in PLN, EUR or USD, net or gross. No surprises on the AI provider's invoice.

The names ChatGPT, Claude, Gemini, Perplexity, DeepSeek and the model names are trademarks of their owners and are used for information only; DualDrive AI is not affiliated with these providers. API charges are collected by the AI provider and are not included in the price of the app. The amounts in the examples were calculated using the app's price list (rates from October 2026, example exchange rate 1 USD ≈ 3.66 PLN); the history, chart, balance and currency rates in the screenshots are sample data.

Want to see the app in action? Let’s talk.

Free consultation