About

Tiny Utility Lab runs Screenshot2Text / SnapOCR as a focused product, not as a side button inside a larger suite. There is no venture-backed growth team rewriting the homepage every week. There is a small operator inbox at hello@screenshot2text.com, published policies, and a converter that tries to stay boring in the good way: drop a file, get text, leave.

Questions about the product, privacy wording, guide corrections, or site feedback go to that same address. The contact page explains what to include and what not to attach. Please do not email screenshots or PDFs. Recognition is local to your device, so we cannot re-run your file on a remote queue even if you send it.

We are not a staffed enterprise support desk. We still read mail and fix clear mistakes when we can.

People pull words out of pictures every day. An error dialog someone pasted into chat. A slide from a meeting. A tracking number in a photo. A scanned form that is still just pixels. Upload converters solve that by shipping the file to a server, running a model, and returning text. That design is common, easy to meter, and fine for some jobs.

It is the wrong default for a lot of other jobs. Work screenshots can include customer names, unpublished designs, internal URLs, or one-time codes. School notes and personal documents are not something most people want sitting on an unknown machine, even briefly, even when a vendor says files are deleted quickly. "Trust us, we delete it" is still a trust decision.

Screenshot2Text uses open-source Tesseract OCR compiled to WebAssembly and run inside the tab you already have open. You choose a language, drop a file, and the page reads pixels on your device. The first run may fetch small support files (the engine, language data, and a PDF page renderer when needed). Those requests ask for named public assets. They do not include your image or the text the tool produced. Your browser caches what it fetched, so later runs are usually quieter.

Closing or reloading the tab clears the working copy in memory. There is no account and no file history on our side, because we never held one.

Saying no keeps the product honest.

We will not add a paywall, a paid button to remove limits, or a subscription that gates basic OCR. The converter is free to use.

We will not require an account to drop a file. Accounts are useful for sync and billing. This tool does not sync your screenshots and does not bill you.

We will not quietly turn local recognition into a mandatory upload pipeline. If the product ever offered an optional cloud path for a specialized model, that would be labeled as a separate choice with a clear privacy cost. Today the product is browser OCR.

We will not promise perfect handwriting, magic cleanup of blurry phone photos of monitors, or courtroom-grade transcripts from decorative fonts. Open-source printed-text OCR has known limits. Overclaiming accuracy is how tools lose trust.

We will not build features whose main job is to help people copy content they are not allowed to use. Acceptable use is covered on Terms.

We will not pretend a tiny utility is an enterprise platform with SLAs, SSO, and custom model training on demand. If you need that stack, you need a different vendor.

Sharp interface screenshots, clear terminal output, printed pages, and well-lit document photos convert most cleanly. Multi-image batches and multi-page PDFs run sequentially in the same reader. English is the default language. Spanish, French, German, Portuguese, Italian, Dutch, Chinese (Simplified), Japanese, and Korean are available from the picker. Matching language data loads with the first read in that language.

Handwriting is weak. Cursive notes, whiteboard markers, and stylized display type often come back with gaps. Heavy JPEG compression, motion blur, glare on a photographed monitor, stamps over text, and faded photocopies also cost accuracy. Very small UI type on a phone screenshot is another common miss. Zoom in the app before you capture, or crop to the region that matters.

Dense multi-column layouts can scramble reading order even when individual words look fine. Cropping to one column or one dialog at a time is slower and usually cleaner. Dark interfaces (light text on a dark panel) are harder for classic OCR pipelines that expect dark ink on a light page. Screenshot2Text flips dark regions before recognition so light-on-dark UI can still read, but the source still has to be sharp.

Local recognition leans on your device. A handful of ordinary screenshots is the intended use. Dozens of huge camera originals may slow or stall the tab because the work is not farmed out to a remote GPU cluster. That tradeoff is the privacy model, not a bug we hide.

After recognition, skim the editable box. Fix an obvious misread, then copy or download. Nothing you type in that box is sent to us.

Every feature is free and stays free. There are no payments, no trials, and no paywall. Screenshot2Text does not charge per file.

Ads load in the page chrome. Ad code does not receive your image, your PDF, or the extracted text. Nothing about your file is used to choose an ad. That separation matters. The converter stays usable without accounts or paywalls.

Cookies, third-party requests, CDN fetches for engine files, and hosting logs are spelled out on Privacy. If an ad unit breaks the drop zone or covers the copy box on your viewport, say so on Contact with the page URL and device details. We care about that because a broken layout is worse than a sparse layout.

Most messy output has a boring cause: too much chrome in the crop, dark low-contrast UI, tiny type, the wrong language selected, or a photo of a screen instead of a real screenshot. The Guides index collects short practical articles on those problems.

Start with:

• How to take a screenshot that OCR reads well
• Fix wrong or messy OCR text
• Private image to text: why browser OCR matters
• Convert a scanned PDF to text in your browser
• OCR for dark mode screenshots
• Batch PDF OCR without upload
• When the tool won't load or stalls

Prefer a real screenshot over a camera shot of a monitor. Prefer PNG when you still have one. Crop status bars, keyboards, and sidebars before you drop the file. Pick the language that matches the text. Run one dense column at a time when layout is messy. Then edit the box like you would edit any rough draft.

• Contact: hello@screenshot2text.com, what to include, what not to attach.
• Privacy: how local recognition, ads, and CDN fetches fit together.
• Terms: acceptable use and accuracy limits.
• Homepage: the free converter itself.

Screenshot2Text / SnapOCR from Tiny Utility Lab is a small tool with a clear job. Read pixels on your device. Give you editable text. Stay free. Stay honest about what OCR cannot do.