Where your data goes.
ToolForte's oldest promise is that the tools run in your browser and your files never reach us. That is true of nearly every tool and it is not true of all of them, and a promise with unmarked exceptions is worth less than a smaller promise you can check. This page is the complete answer, including the parts that are less flattering to us.
The four answers
Every tool page tells you which of these applies to it, before you type anything in. This is what each one means.
This runs in your browser. What you put in is never sent to ToolForte, so we could not read it if we wanted to.
Close the tab and it is gone.
The ordinary case, and nearly all of the 341 tools. Nothing is uploaded, so there is nothing for us to keep, lose or hand over.
This is saved by your browser on this device. It is not sent to ToolForte, so it is not on your other devices and we cannot recover it for you.
Clearing your browser data removes it, and so does the delete button here.
30 tool pages, the ones that remember your work between visits: planners, the invoice and resume builders, a few games keeping a score.
Your browser asks an outside service for this directly. It does not pass through ToolForte, so we never see what you looked up, but that service does, and its own privacy policy governs what it keeps.
Nothing for us to remove, because we never received it.
5 tool pages, where your browser asks somebody else's service directly. Each one names the service on the page.
This step runs on ToolForte servers, so the input is sent to us to be processed. It is not stored afterwards; only the fact that the call happened is recorded, for your usage history.
Nothing to remove: the content is discarded once the response is returned.
8 tool pages, plus every REST API call, MCP tool call and render.
The tools that send what you type to us
Four of these check a website you name, so our server has to be the one to fetch it. Four are the AI writing tools.
In every case what you send is processed and not kept. What remains afterwards is the record that a call happened, which is what your usage history and your credit balance are made of, and it does not include what was in the call.
- ai-blog-outline: Your topic is sent to our server and on to the model provider that writes the outline.
- ai-email-writer: What you describe is sent to our server and on to the model provider that writes the email.
- ai-review-response: The review you paste is sent to our server and on to the model provider that drafts the reply.
- ai-tone-changer: The text you paste is sent to our server and on to the model provider that rewrites it.
- broken-link-checker: Our server fetches the page at the address you entered and checks the links on it.
- canonical-checker: Our server fetches the page at the address you entered and reads its canonical tag.
- http-headers-checker: Our server requests the address you entered and reports the headers it got back.
- meta-tag-checker: Our server fetches the page at the address you entered and reads its meta tags.
The AI tools have a second half worth stating plainly. Your text goes to our server and on to the model provider that generates the answer, so a company other than us sees it. If that is not a trade you want to make for a given piece of text, the honest advice is not to paste it.
The tools that ask somebody else directly
These never touch our servers at all, which sounds better than it is.
"Not us" is not the same promise as "nobody". Your browser sends the domain or address you typed straight to an outside company. We genuinely cannot see it, and we equally cannot tell you what that company keeps, so each page names the service and links to its own policy rather than reassuring you on its behalf.
- currency-converter: Frankfurter
- dns-lookup: Google Public DNS
- domain-checker: Google Public DNS
- ip-address-lookup: ipify and ipapi
- whois-lookup: rdap.org
The render API, which is the one that stores a file
Everything above either keeps your data on your device or discards it. This is the exception.
The render endpoints and their MCP tools run headless Chrome on our servers to produce a PDF or a screenshot. What you send is processed there, and the file that comes out is written to a private bucket and handed back as a signed link that expires after 24 hours. A daily job deletes the files themselves. The bucket has no public access, so a file is reachable only by someone holding the signed link, which means treating that link as the file.
An account, and our own visit counting
Page views
We count page views and tool openings ourselves rather than leaving it entirely to Google. What is recorded is the path, which tool it was, a random session id, and the referring site's host when it was not us. Not an IP address, not a raw browser string, and nothing you typed into a tool. If you decline cookies, nothing is written to your device at all: the id lives in memory for that one page load. The privacy policy has the full detail.
Workflow runs that happen on your device
A workflow whose every step is pure computation runs in your browser, and its input and results never reach us. We do record that the run happened: which ready-made workflow it came from if it came from one, how many steps it had, whether it finished, and how long it took. Never what you put in, never what came back, never anything written into the steps, and never the workflow's name or id.
It is the one thing a browser run sends, and we would rather name the four fields than call it anonymous usage data, which is a category rather than a list. The sentence also sits next to the run button, because the place to learn this is where the decision is made.