Privacy

Optical character recognition runs inside your browser tab using open-source Tesseract software. When you drop a PNG, JPG, WebP, GIF, BMP, or PDF, the page reads the pixels locally and writes the result into a text box you can edit and copy. PDF pages are drawn in the same tab before they are read. We do not upload your file for recognition, and there is no server behind this site doing the reading for you.

We do not receive your image. We do not receive your PDF. We do not receive the text the tool extracts. Those stay in the browser tab while the page is open and are gone when you close or reload it.

There is no account, no sign-in, and no file history. We cannot sell or share your OCR content, because we never hold a copy of it.

The first recognition run may fetch small support files to your device: the OCR engine, the language data for the language you picked, and a page-rendering script when you open a PDF. These come from public content delivery networks so the work can happen locally.

Those requests ask for a named file. They do not include your image, your PDF, or the extracted text. Your browser caches what it fetched, so later reads usually skip the download.

The site is free and paid for by advertising through Google AdSense. Google and its partners may set or read cookies, use device identifiers, and log standard request data such as IP address, browser type, and the pages viewed here in order to serve and measure ads. Depending on your region and your settings, some of that may be used for personalized advertising, and a consent choice may appear before it is.

Live ad units sit in the page chrome on the homepage converter, not inside the drop zone or the editable result box. That keeps recognition usable on phones and laptops. Ads never fund a private copy of your file on our side, because recognition never sends one.

Ads are separate from OCR. Ad code does not receive your image, your PDF, or the text the tool extracted. You can review how Google uses data from sites that show its ads at policies.google.com/technologies/partner-sites, and change your ad settings at myadcenter.google.com.

We do not set our own tracking cookies. Advertising vendors may set cookies as described above.

The site keeps one small preference in your browser: whether you chose the light, dark, or system theme. It is stored in local storage on your device, it is not sent to us, and clearing site data removes it.

The site is hosted on Vercel. Like most hosts, Vercel processes ordinary request information such as IP address, user agent, and the URL requested in order to deliver pages and keep the service running. That is normal web traffic. It does not include your file or your extracted text, because neither one leaves the tab.

Three outside services are involved: Google AdSense for advertising, Vercel for hosting, and public content delivery networks for the OCR and PDF files described above. Each runs under its own privacy policy, and their handling of the data they receive is governed by those policies rather than this page.

Screenshot2Text is a general-purpose utility. It is not directed at children under 13, and we do not knowingly collect personal information from children. There is no account to create and no profile to fill in.

Product questions, privacy notes, and site feedback go to hello@screenshot2text.com. That is the only support inbox. There is no account mailbox and no in-app ticket store.

When you write, we receive the contents of the message: your From address, the subject, the body, and any headers the mail system includes. A human at Tiny Utility Lab reads that mail to answer or to fix the site. We do not run recognition on attachments, and we do not need your screenshot, photo, PDF, or extracted text to investigate most bugs.

Please do not attach files. Emailing a screenshot or PDF creates a copy outside your browser, which undercuts the local-OCR design. Describe the page URL, what went wrong, the browser and device, the language you selected, and whether the failure happened on an image or a PDF. That written detail is enough. If someone still attaches a file, we delete the attachment when we can and reply from the text of the message alone. Do not send secrets, passwords, one-time codes, or personal documents "just in case."

We keep ordinary correspondence only as long as it is useful for the reply or a follow-up fix, then clear it from the active inbox. We do not build a marketing list from contact mail, and we do not sell addresses. For response timing and what we can or cannot help with, see the contact page.

You can manage cookies and site data in your browser settings and opt out of personalized ads through Google. You can use the tool without an account. Because recognition runs on your device, there is no stored file or saved transcript for us to hand over or delete on request.

Clearing site data for screenshot2text.com removes the theme preference and any cached engine files your browser kept. The next recognition run may download those public support files again. Blocking third-party cookies or using a stricter tracking-protection mode can change which ads appear; it does not upload your image or give us a copy of the text in the result box.

If you emailed hello@screenshot2text.com and want that thread removed from the active inbox after we have replied, say so in a short follow-up. We cannot delete a file we never received for OCR, because that file never left your tab.

We update this page when the tool or the services behind it change. The date at the top shows the last revision. Using the site after a change means the current version applies.

Screenshot2Text, also known as SnapOCR, is operated by Tiny Utility Lab, a small independent maker of free browser utilities. For privacy questions or other site feedback, email hello@screenshot2text.com. The same address is listed on the contact and about pages.

The same commitments in plain form are on the terms page.