Guide
What is the image-to-text tool?
This is a free image-to-text converter (OCR) that reads the text out of screenshots, scans, and photos. The OCR engine (PaddleOCR compiled to WebAssembly) runs entirely in your browser, so the image is never uploaded and never leaves your device. It is free with no signup, and the extracted text can be copied or downloaded as a .txt file.
Here is what it reads well and what it does not:
| Input | Fit | Notes |
|---|---|---|
| Screenshot with text | Best | Crisp pixels, high contrast: ideal OCR input |
| Scan or straight-on photo of a printed page | Good | Flat, well-lit pages work best |
| Photo of a sign, label, or package | Good | Get close enough that letters are sharp |
| Handwriting, whiteboard | Poor | Use Transcribe an image with AI instead |
| Tables, multi-column layout | Text only | Reading order may flatten; the AI tool reconstructs structure as Markdown |
When it comes in handy
- All you have is a screenshot of an error message: turn it into text and paste it straight into a search engine or a support ticket.
- Quoting part of a paper document: transcribe just the passage you need instead of retyping it.
- Jotting down the text on a sign, label, or package: shoot it with your phone and get text on the spot.
- Reusing material that only survives as an image.
- Running OCR on confidential documents without uploading them anywhere: everything stays on your device.
How to use
- Drop an image onto the tool. On a phone, tap "Take a photo" to shoot it right there; clicking to choose a file also works. PNG, JPEG, and WebP are supported.
- Wait a moment. The first run downloads the OCR model (~25MB), which is cached afterwards.
- Read the extracted text. Lines come out in reading order as plain text.
- Copy the text or download it as a .txt file.
For example, a screenshot of an error dialog reading "The operation couldn't be completed. Error code 0x80070005" returns exactly that text, ready to paste anywhere.
Notes
- Flat, well-lit, straight-on images give the best accuracy. Strong perspective, blur, or glare reduces it.
- Output is plain text only: no headings, no tables, no formatting.
- For handwriting, or when you want the structure preserved as Markdown, use Transcribe an image with AI. For handwritten notes specifically, there is Transcribe handwriting with AI.
