Where your files go: browser tools and AI tools
Last reviewed: 9 October 2026
Some of our tools run entirely in your browser; others send what you give them to a model. That difference matters more than any promise we could make, so this page sets out where the line falls, what we receive either way, and how to verify it without taking our word for it.
Two kinds of tools
Browser tools do the work on your device: the file is read by the page you already have open, and the result is produced locally. AI tools send the input you provide to our API, where a model processes it and returns the result. Each tool page states which kind it is before you run anything.
What a browser tool means
Three things are true when a tool runs in your browser.
Your file stays on your device
The file is opened by your browser and never sent to us. There is no upload step, and no copy of it reaches our servers.
You still download our code
The page and its tool code are served by us, like any website. Local means your file is not processed on our side — not that nothing at all is requested.
The result is in your hands straight away
Nothing has to round-trip to a server, so the export happens as soon as your browser finishes the work.
What an AI tool sends
These tools need a model, so the input has to travel.
The input you submit goes to us
For an AI tool, the text or file you provide is sent to our API so a model can process it. That input and the result are stored with your task history.
We pass it to a model through our API
Processing happens as a model call from our API. Keeping the prompt and the result in your task record is what lets you review past runs and re-run a job.
AI runs are priced in credits
Each AI run costs credits according to its pricing rule, shown on the tool page before you start.
What we receive either way
Running your files locally does not mean we know nothing. This is what we still hold.
| Item | What we hold |
|---|---|
| Account | Your email address, and a hashed password if you sign up with email. We store a hash, not the password itself. |
| Usage and credits | Which tool ran, when, and how many credits it used — for browser tools as well as AI tools. Browser tools still cost credits, so the balance has to be recorded. |
| Analytics — only with your consent | We measure which pages and tools are used, but only after you accept. Before that no analytics request is sent, and if you withdraw consent we stay silent. |
| Payments | Handled by a third-party payment provider, so card details never reach us. |
Common misreadings
Four things people reasonably assume that are not quite right.
| The assumption | What is actually true |
|---|---|
| Local means it works offline | It does not. The page and the tool code are served by us; local refers to where your file is processed, not to whether the site is reachable. |
| Nothing is uploaded, so you know nothing about me | We still hold your account details and a record of which tools ran and when. The page request itself also reaches our servers, as it would on any site. |
| Browser tools are free | They follow the same pricing rules and cost credits per run, which members do not pay. The price is shown on the tool page. |
| AI tools keep copies of my files | AI tools store the input you submitted and the result — that is what your task history is. Browser tools leave us with nothing of the file to store. |
Check it yourself
You do not have to take any of this on trust. A few minutes with your browser's developer tools will show you the difference.
- Open the network panel
In your browser, open the developer tools and switch to the Network tab.
- Run a browser tool
Pick one of the tools below, add a file, and run it while you watch the panel.
- Look for uploads
You will see the page's own assets load. What you will not see is a request carrying your file off your machine.
- Run an AI tool and compare
Now run an AI tool on the same page and watch the panel again: you will see the request that carries your input to our API, and the calls that fetch the result.
- Read the policy
The binding account of what we store, how long we keep it and how to have it deleted is the privacy policy, linked below.
Common questions
Are browser tools really offline?
No. The page and its code come from our servers like any website. What stays local is your file: it is not uploaded and not processed on our side.
Can you see the files I use in a browser tool?
Not through the tool. The file is read by your browser and the result is produced there, so there is no request that would carry it to us.
When does analytics start?
Only after you accept. Before that no analytics script is loaded and no request is sent; if you withdraw consent, reporting stops.
Why do browser tools cost credits if nothing runs on your servers?
Credits are how the service is paid for. The pricing rule applies per tool, and members do not pay for browser tools. The price is always shown before you run anything.
Do you keep what I send to an AI tool?
The input and the result are stored with your task history, which is what lets you review past runs. The retention rules and how to have it deleted are in the privacy policy.
Try it yourself
These run in your browser — open the network panel first and watch what does, and does not, leave your machine.
Where this is documented
This page describes how our tools behave today. The binding version of what we collect, why, how long we keep it and how to have it deleted is the privacy policy.
If this page and the privacy policy ever disagree, the policy is the one that applies — and we fix this page.