Image to Text
Copy printed words from a screenshot, photo or scanned image without typing them again. Choose the text language, run OCR and correct the result against the original before using it.
Before you start
- Input
- One printed-text JPG, PNG or static WebP; select its language.
- Output
- Editable plain text; OCR is a preview feature.
Loading tool…
How to use this tool
- Choose a supported image and select its text language.
- Start recognition and wait for the language resources and processing to finish.
- Review and correct the text, then copy it or save a TXT file.
Understand the result
OCR estimates characters from image pixels. It can confuse similar letters, digits and punctuation, especially in small or blurred text. The output is editable plain text; it is not a recreation of the original document layout.
How do I copy text from a screenshot or photo?
Choose the original image and select the language printed in it. Run recognition, compare the result with the picture and correct names, numbers and punctuation before copying or downloading TXT. For several columns, cropping to one text block can make the output easier to review.
Can I turn a scanned PDF into editable text here?
This tool accepts pictures, not PDF files. First export the relevant PDF page as JPG, then choose that image for OCR. You receive plain text, not a recreated Word layout or a searchable PDF. Recognition remains a preview feature; Traditional Chinese and Urdu may need substantial correction.
Try a sample
Try an original sample to inspect the supported workflow. A successful sample is not a guarantee for every input.
Original image sample ↓
If the result is not what you expected
Words, numbers or columns are incorrect.
Use a sharp, upright original and the correct recognition language. Crop to one text block when columns mix. Compare names and numbers against the image; Traditional Chinese and Urdu may need substantial correction.
Next useful tool: Crop Image →Important limits
- One image at a time; PDF files need a different workflow.
- Recognition quality varies by image and language. Check names, figures and punctuation.
- Plain text output does not preserve columns, fonts or table structure.
- One JPG, PNG or static WebP up to 10 MB and 12 million pixels. One recognition language per run.
- OCR preview: recognition varies by font and script. Traditional Chinese and Urdu may need substantial manual correction. Always review the result.
Frequently asked questions
Are my pictures sent to a conversion server?
The image is processed in your browser. Recognition resources are downloaded when needed, but the image and recognized text are not uploaded.
Why does the first run take longer?
The browser may need to download and prepare the selected language resources before recognition. Later runs can reuse cached resources.
Can it recognize handwriting or preserve tables?
This release is designed for printed text. Handwriting, mathematical notation, complex tables and difficult photographs may produce poor results.
Can I use a different language from the website interface?
Yes. Choose the language present in the image. The language of the controls does not necessarily match the text you want to recognize.
Related guides
Questions or a correction? mrkcal@mrkcal.com