Image to Text Converter — Free OCR, No Upload
Copy the text from a photo, screenshot or scan in seconds. OCR runs in your browser, so images are never uploaded. 13 languages, including Hindi and Bengali.
Loading only the code needed for this tool…
Accuracy we measured
We ran this tool on our own test images (no customer files) and compared its output with the known text. The character error rate counts wrong, missing and extra characters; 0% means every character was right.
| Test image | Language | Errors | Engine defaults | Time | Phone profile |
|---|---|---|---|---|---|
| Same notice, phone photo at an angle | English | 0.1% | 2.4% | 3.7 sec | 6.2 sec |
| App screenshot, 15 px text | English | 1.6% | 3.2% | 1.8 sec | 6.5 sec |
| Café receipt, phone photo | English | 0% | 1.6% | 1.9 sec | 4 sec |
| Two-column newsletter | English | 0% | 58% | 5.8 sec | 11.3 sec |
| Same photo in a dim room | English | 1.5% | 3.3% | 4.8 sec | 5.4 sec |
| Hindi page, phone photo | Hindi | 0% | 0.4% | 4.1 sec | 7.3 sec |
| Bengali page, phone photo | Bengali | 0% | 0.2% | 1.9 sec | 4.9 sec |
| Simplified Chinese page, phone photo | Chinese (Simplified) | 1.4% | 23.4% | 1.7 sec | 5.7 sec |
| Japanese page, phone photo | Japanese | 2.6% | 25% | 1.6 sec | 4.1 sec |
| Printed notice, flat scan | English | 0% | 0% | 5.9 sec | 6.7 sec |
| Printed Hindi page | Hindi | 0% | 0% | 1.8 sec | 4.3 sec |
| Printed Bengali page | Bengali | 0% | 0% | 1.8 sec | 4.1 sec |
| Printed Japanese page | Japanese | 1.5% | 1.5% | 1.9 sec | 3.8 sec |
| Printed Simplified Chinese page | Chinese (Simplified) | 0% | 0% | 1.8 sec | 4.3 sec |
| Printed Traditional Chinese page | Chinese (Traditional) | 3.6% | 3.6% | 1.8 sec | 4.4 sec |
| Printed Korean page | Korean | 0.4% | 0.4% | 1.8 sec | 3.3 sec |
| Printed Arabic page | Arabic | 0% | 0% | 1.4 sec | 4.9 sec |
| Printed Spanish page | Spanish | 0% | 0% | 3.2 sec | 7.6 sec |
| Printed French page | French | 0% | 0% | 1.1 sec | 3.5 sec |
| Printed German page | German | 0% | 0% | 2.9 sec | 4.4 sec |
| Printed Portuguese page | Portuguese | 0% | 0% | 2.6 sec | 5.7 sec |
| Printed Italian page | Italian | 0% | 0% | 2.7 sec | 4 sec |
Engine defaults: the same engine and language data without our clean-up (automatic layout, no straightening, no light correction).
How we measured: the images come from a script that renders our own sentences, photographs them at a 4° angle with uneven light, and saves JPEGs, PNGs and PDFs. Times are from a 4-core cloud computer; the phone profile adds Chrome's 4× CPU slowdown. First runs also download the engine (about 1.1 MB) and the language data.
Clean printed pages in all 13 languages we offer came out with 3.6% or fewer errors. Photos of Chinese and Japanese text are harder: keep the camera straight above the page.
Tool quality check
Use the output as a verified working copy
Keep the original source when it may be needed again. Open the generated TXT in the application that will receive it, and check that its content and formatting meet your needs.
Input
.jpg, .jpeg, .png, .webp
Output
TXT
Batch
Up to 20 files per run
Up to 20 files per run. Only use one run when files share the same operation.
Move into a related operation instead of re-finding the tool.
Download a destination-ready result and verify it before sharing.
Configured runtime limit applies.
JPG, JPEG, PNG, WEBP, HEIC, HEIF, BMP, GIF
TXT
Check the upload limit below
Review the result before sharing
Smart workflows
Finish the job, not just the file
Use the existing tool graph to continue into useful next steps. Bulk and output-packaging capabilities are surfaced automatically where the underlying tool supports them.
Continue with Jpg To Word
Bulk-readyFinish the job after Image To Text with a relevant MAXScanner tool, without creating a duplicate converter page.
- Image To Text
- Jpg To WordBulk processing · ZIP-ready output · Quality controls
Best when Image To Text is only the first step in your workflow.
Continue with Pdf To Text
Bulk-readyFinish the job after Image To Text with a relevant MAXScanner tool, without creating a duplicate converter page.
- Image To Text
- Pdf To TextBulk processing · ZIP-ready output · Page selection
Best when Image To Text is only the first step in your workflow.
About This Tool
To turn an image into text, add the photo, screenshot or scan, choose the language of the text and select Extract text. The words appear in an editor a few seconds later, ready to copy or to download as TXT, Word (.docx) or a searchable PDF. Recognition runs inside this browser tab with the Tesseract engine, so the image is never uploaded, and there is no sign-up or daily limit.
JPG, PNG, WebP, HEIC (iPhone photos), BMP and GIF all work in this one tool, so there is no separate JPG to text or PNG to text page. Add up to 20 images at once, take a photo with your phone's camera, or paste a screenshot with Ctrl+V.
What happens before the text is read
Most mistakes in free OCR come from the photo rather than the engine. Before Tesseract reads an image, MAXScanner straightens a tilted photo, evens out shadows and dim light, enlarges small screenshot text and checks whether the page has two columns. On our phone photo of a printed notice, taken at an angle, this took the character error rate from 2.4% with the engine's default settings to 0.1%. On a photographed Chinese page it went from 23% to 1.4%, and on a two-column newsletter from 58% to 0%, because the default settings read straight across both columns, line by line.
When this is not the right tool
- Handwriting. The engine is trained on printed text; handwritten notes come out mostly wrong.
- Tables. You get the text line by line, not rows and cells. For tables in a PDF, try Extract tables.
- Keeping the page as it looks. To keep a scan as a document with the page image and selectable text, use OCR PDF. For an editable Word file of a photographed page, use JPG to Word.
- Badly blurred or very dark photos. Straightening and light correction help, but they cannot restore letters the camera never captured. Take the photo again, closer, with the page flat and in good light.
After extracting: put the text into a PDF with TXT to PDF, combine the photos themselves into one PDF with Image to PDF, or get the text out of a PDF with PDF to Text.
How to Use
Add the image
Drop or choose JPG, PNG, WebP or HEIC files (up to 20), take a photo with your phone, or paste a screenshot with Ctrl+V.
Choose the language
Pick the language of the text, or up to three such as Hindi and English. Leave Page layout on Automatic unless the page has columns that are read in the wrong order.
Extract, check and copy
Select Extract text. Correct any mistakes in the editor, then copy the text or download it as TXT, Word or a searchable PDF.
Use Cases
Screenshots and messages
Copy an address, order number or error message out of a screenshot instead of retyping it. Small 1x screenshot text is enlarged before it is read.
Printed notices and handouts
Photograph a notice, recipe or worksheet and paste the words into your notes, an email or a document.
Hindi, Bengali and other scripts
Read printed Hindi, Bengali, Arabic, Chinese, Japanese and Korean as well as English and other Latin-alphabet languages, or a mix of two or three.
Frequently Asked Questions
How do I copy text from an image?+
Add the image on this page (or paste a screenshot with Ctrl+V), choose the language of the text and select Extract text. The text appears in an editor where you can fix mistakes and select Copy all.
Is this image to text converter free? Are there limits?+
It is free, with no sign-up and no daily limit. You can add up to 20 images per batch. Because the work happens on your device, large phone photos take longer on older phones.
Are my images uploaded to a server?+
No. The text is recognized in your browser. On first use your browser downloads the recognition engine and the language data from the jsDelivr CDN and keeps a copy; your images and the text stay on your device.
Which languages can it read?+
13: English, Hindi, Bengali, Spanish, French, German, Portuguese, Italian, Japanese, Chinese (Simplified and Traditional), Korean and Arabic. In our test, clean printed pages in all of them came out with 3.6% character errors or fewer. You can combine up to three, for example Hindi and English.
Can it convert handwriting to text?+
No. The engine is trained on printed text, so handwriting gives poor results. Use it for printed pages, screenshots, receipts and signs.
Why does the extracted text have mistakes?+
Usually because of the photo: blur, glare, deep shadows, tiny text or the wrong language setting. Take the photo straight above the page in good light and check the language. Results with low confidence are flagged so you know to read them carefully.
How do I convert a JPG or PNG to text?+
The same way: this tool accepts JPG, PNG, WebP, HEIC, BMP and GIF. PNG stores screenshots without compression loss, so very small text stays sharper than in a heavily compressed JPG.
Can I get a Word file or a PDF instead of plain text?+
Yes. After extracting, download Word (.docx) or a searchable PDF that shows your image with an invisible text layer. For a Word file as the main result, JPG to Word does the same with Word as the default download.