Go to affligo.com/ocr-pdf
Drag & drop or click to browse
Select language, click Run OCR
Get your searchable PDF
A scanned PDF is just an image of text. You cannot search it, copy from it, or highlight it. Whether it is a contract scanned at the office, a research paper digitized from a library, or a receipt photographed on your phone — the text is trapped inside a picture, invisible to computers.
Optical Character Recognition (OCR) solves this problem. It analyzes the image, identifies characters and words, and creates a hidden text layer underneath the visual content. The result is a PDF that looks identical but now supports full-text search, copy-paste, screen readers, and digital indexing.
This guide covers everything you need to know about OCR for PDFs: how it works under the hood, what languages are supported, how to configure the processing pipeline for best results, and how to use AFFLIGO's free online OCR tool to convert any scanned PDF into a fully searchable document in seconds.
Video Walkthrough Coming Soon
A complete step-by-step demo of OCR processing scanned PDFs with language selection and output configuration.
Why Scanned PDFs Need OCR
Without OCR, a scanned PDF is just a picture of text. Here is why that matters.
No Text Search
A scanned PDF contains zero searchable text. Ctrl+F does nothing. You cannot find a single word, name, date, or number. This makes referencing and auditing documents painfully slow — you have to read every page manually to locate specific information.
No Copy-Paste
Try selecting text from a scanned PDF and you get nothing — or worse, you select the entire page as one image. Without OCR, you cannot copy quotes, extract data, or paste text into another application. You must retype everything manually, which is slow and error-prone.
Not Accessible
Screen readers and assistive technologies cannot read scanned PDFs. The text is embedded as pixels, not characters. This makes your documents inaccessible to visually impaired users and violates accessibility compliance standards like WCAG 2.1 and Section 508.
Cannot Be Indexed
Search engines, document management systems, and digital archives cannot index content inside scanned images. Your documents become invisible to search. Without OCR, a thousand-page scanned archive is essentially a black box — impossible to search, sort, or filter.
Collaboration Blocked
When you share a scanned PDF with colleagues, they cannot search for key terms, annotate specific text passages, or extract data into spreadsheets. The document becomes a static image that everyone must read linearly, slowing down review cycles and team productivity.
Legal & Compliance Risk
In legal discovery and regulatory compliance, you must produce searchable documents. Courts and regulators expect electronic records that can be searched, reviewed, and produced on demand. Sending scanned images without OCR can result in sanctions, penalties, and adverse inferences.
OCR transforms a static image into a living, searchable document. The moment you run OCR on a scanned PDF, it becomes as functional as a digitally-born document — searchable, copyable, accessible, and indexable. For any document you plan to reference, share, or archive, OCR is not optional — it is essential.
What Is OCR (Optical Character Recognition)
The technology that turns images of text into real, machine-readable characters.
Optical Character Recognition (OCR) is a technology that identifies and extracts text from images — photographs, scans, screenshots, and PDFs containing image-based content. OCR analyzes the visual patterns of characters, matches them against trained models, and produces machine-readable text output. For scanned PDFs, this means creating a hidden text layer that sits underneath the visual image, enabling search, copy-paste, and accessibility.
Image Preprocessing
Before OCR can read the text, the image must be cleaned and prepared. This step includes:
- Deskew correction — straightens tilted or rotated scans so text lines are horizontal
- Noise removal — eliminates speckles, dots, and background artifacts from the scan
- Contrast enhancement — improves the distinction between text and background
- Binarization — converts the image to pure black and white for clearer character boundaries
Good preprocessing is the foundation of accurate OCR. A clean, straight image with high contrast produces dramatically better results than a blurry, tilted scan.
Text Line Detection
The OCR engine identifies where text exists in the image. It uses algorithms to detect:
- Text lines — horizontal rows of characters
- Word boundaries — gaps between words
- Character regions — individual character shapes
- Layout structure — columns, headers, footers, tables, and margins
Modern OCR engines like Tesseract use connected-component analysis and projection profiles to segment text regions from images, even in complex multi-column layouts.
Character Recognition
Once text regions are identified, the engine recognizes individual characters using:
- Pattern matching — comparing character shapes against trained font models
- Neural networks — deep learning models trained on millions of character samples
- Language models — using word probability and context to resolve ambiguous characters
- Dictionary lookup — validating recognized words against a language dictionary
Tesseract uses Long Short-Term Memory (LSTM) neural networks, which understand character sequences in context — recognizing "rn" as "m" when it appears within a known word.
Text Layer Creation
The final step creates the output. For PDF OCR, this means:
- hOCR format — an XML-based format that stores character positions, confidence scores, and bounding boxes
- PDF embedding — the recognized text is placed as an invisible layer underneath the original image
- Searchable result — the output PDF looks identical but now supports full-text search, copy-paste, and screen reader access
The original image is preserved exactly as it was. OCR adds a text layer without modifying the visual content, maintaining document fidelity.
OCR accuracy depends heavily on scan quality. A 300 DPI scan with clean text and good contrast can achieve 98-99% accuracy. A low-resolution phone photo with shadows and blur may drop to 85-90%. Always scan at 300 DPI or higher, ensure good lighting, and keep the document flat against the scanner glass for best results.
How Server-Side OCR Works
The technical pipeline from upload to searchable PDF — Stirling PDF, Tesseract, and the processing chain.
AFFLIGO's OCR tool uses server-side processing via Stirling PDF, an open-source PDF manipulation suite that integrates Tesseract OCR — the most widely used open-source OCR engine. Here is what happens when you click "Run OCR":
Secure Upload
Your PDF file is uploaded to the server via an encrypted HTTPS connection. The file is sent as multipart form data along with your configuration parameters — language, OCR type, render type, output type, and processing options. The upload uses XHR with progress tracking so you see real-time status.
Server-Side Processing via Stirling PDF
Stirling PDF receives the file and routes it to the OCR processing pipeline. The server extracts pages from the PDF, converts each page to an image format suitable for OCR, and passes them to Tesseract. Stirling handles the PDF structure, metadata, and page assembly — Tesseract focuses purely on character recognition.
Tesseract OCR Engine
Tesseract processes each page image using its LSTM neural network. It detects text lines, segments characters, and recognizes them using language-specific models. The engine supports 38 languages and can process multiple languages simultaneously. Each character receives a confidence score, and the engine uses language models to resolve ambiguous characters.
hOCR Generation
Tesseract produces output in hOCR format — an XML-based standard that stores recognized text along with precise bounding box coordinates, font information, and confidence scores for every word and character. This format preserves the spatial layout of the original document, ensuring the text layer aligns correctly with the visual content.
PDF Assembly & Download
Stirling PDF takes the hOCR output and the original PDF, merges them into a single searchable PDF document, and returns the result. The server sends the processed file back as a binary response. Your browser receives the blob, and you can download it with a single click. The output file is named with an "-ocr" suffix for easy identification.
Server-side processing means you do not need powerful hardware or installed software. Tesseract runs on the server, so even a phone or tablet can process a 50MB scanned PDF. The trade-off is that files must be uploaded — for documents that cannot leave your device, consider installing Tesseract locally or using a desktop OCR application.
Supported Languages
AFFLIGO's OCR tool supports 38 languages via Tesseract's trained models. Enter the language code(s) in the Languages field.
| Language | Code | Language | Code |
|---|---|---|---|
| English | eng | Lithuanian | lit |
| Spanish | spa | Latvian | lav |
| French | fra | Estonian | est |
| German | deu | Catalan | cat |
| Chinese (Simplified) | zho | Hebrew | heb |
| Japanese | jpn | Ukrainian | ukr |
| Korean | kor | Bulgarian | bul |
| Arabic | ara | Serbian | srp |
| Portuguese | por | Malay | msl |
| Italian | ita | Croatian | hrv |
| Dutch | nld | Slovak | slk |
| Russian | rus | Slovenian | slv |
| Polish | pol | Indonesian | ind |
| Turkish | tur | Hungarian | hun |
| Greek | ell | Czech | ces |
| Thai | tha | Romanian | ron |
| Vietnamese | vie | Finnish | fin |
| Swedish | swe | Danish | dan |
| Norwegian | nor |
For multilingual documents, enter multiple language codes separated by plus signs — for example, eng+fra for English-French documents or eng+deu+fra for trilingual content. Tesseract will load all specified language models and use them during recognition. Adding more languages increases processing time but improves accuracy for mixed-language documents.
Step-by-Step: How to Use AFFLIGO OCR
Seven detailed steps from upload to downloadable searchable PDF.
Upload Your Scanned PDF
Open affligo.com/ocr-pdf in your browser. Drag and drop your scanned PDF onto the upload area, or click to browse and select the file from your device.
- File format: PDF only (.pdf extension required)
- Maximum size: 50MB per file
- File naming: The output will automatically append "-ocr" to your filename
- Multiple files: You can upload and process one file at a time
The tool will display your file name and size in the file list. Click the X button to remove a file and start over if needed.
Set the Document Language
In the Languages field, enter the Tesseract language code for your document. The default is eng (English).
- Single language: Enter one code, e.g., eng
- Multiple languages: Use plus signs, e.g., eng+spa or eng+fra+deu
- Unknown language: Use eng as a fallback — it handles most Latin scripts acceptably
- CJK languages: Use zho (Chinese), jpn (Japanese), or kor (Korean) for East Asian text
If you are unsure of the language, start with eng and reprocess with the correct language code if accuracy is low.
Choose Your OCR Type
Select how the OCR engine processes your document. This controls what gets recognized and what is skipped.
- Force OCR (default): Processes every page through OCR regardless of existing text. Best for fully scanned documents.
- Skip All Text: Skips pages that already contain a text layer. Processes only image-only pages.
- Skip Text & Meta: Skips existing text layers and metadata. Useful for mixed documents.
- Only OCR: Processes only pages that need OCR. Skips already-searchable pages.
- Skip OCR: Passes the document through without OCR. Useful for metadata cleanup only.
For most scanned PDFs, Force OCR is the correct choice. It ensures every page gets processed regardless of what the server detects.
Pick Render Type and Output Type
Configure how the OCR output is rendered and what format it produces.
- hOCR (default): Standard hOCR rendering. Produces clean, well-structured searchable PDFs. Recommended for most use cases.
- hOCR Original: Preserves the original page layout more closely. Use when the standard rendering distills formatting or spacing.
- PDF (default): Standard PDF output. Compatible with all PDF viewers and readers.
- PDF/A: Archival PDF format designed for long-term preservation. Required by many legal and government agencies for document archival.
Use hOCR + PDF for everyday use. Switch to hOCR Original + PDF/A for legal, archival, or compliance-critical documents.
Adjust Processing Options
Toggle four checkboxes to fine-tune the OCR pipeline:
- Auto-deskew pages (default: ON): Automatically straightens tilted or rotated pages before OCR. Keep this enabled — it dramatically improves accuracy for imperfect scans.
- Clean image before OCR: Removes noise, speckles, and artifacts from the scan before recognition. Useful for low-quality scans with visible noise.
- Clean image after OCR: Post-processes the output to clean up artifacts introduced during recognition. Can improve the visual quality of the output.
- Remove images after processing: Strips embedded images from the output PDF, leaving only the text layer. Reduces file size but removes visual content.
For most documents, keep Auto-deskew ON and the other three options OFF. Only enable cleaning or image removal when you have a specific reason.
Run OCR
Click the "Run OCR" button. The tool will:
- Upload your file to the server with a progress indicator showing percentage and megabytes transferred
- Process the document through the Stirling PDF + Tesseract pipeline
- Display a real-time progress ring showing completion status
- Show "OCR complete!" when processing finishes
Processing time depends on file size and page count. A typical 10-page document takes 5-15 seconds. Larger documents with 50+ pages may take 30-60 seconds.
Download Your Searchable PDF
Once processing is complete, click the "Download" button. Your browser will save the file with the "-ocr" suffix appended to the original filename.
- Verify the result: Open the downloaded PDF and try Ctrl+F to search for text
- Test copy-paste: Select text and paste it into a text editor to confirm it works
- Check accuracy: Spot-check a few pages to ensure OCR recognized characters correctly
- Reprocess if needed: If accuracy is low, try a different language code or enable the "Clean image before OCR" option
The original image is preserved in the output. You get a PDF that looks identical but now has a fully functional text layer underneath.
For documents with mixed content — some pages scanned, some digitally born — use the "Skip All Text" OCR type. This processes only the image-only pages while leaving already-searchable pages untouched, saving processing time and avoiding double-processing artifacts.
OCR Types Explained
When to use each of the five OCR type options and what they do to your document.
| OCR Type | What It Does | Best For |
|---|---|---|
| Force OCR | Processes every single page through OCR regardless of whether a text layer already exists. Overwrites any existing text recognition. | Documents that are entirely scanned, or when you want a fresh OCR pass on every page for maximum accuracy. |
| Skip All Text | Checks each page for existing text. If a text layer is found, the page is skipped. Only image-only pages get processed. | Mixed documents where some pages are already searchable and others are scanned images. Saves processing time. |
| Skip Text & Meta | Skips existing text layers and document metadata. Focuses OCR on image content while preserving existing metadata. | Documents with valuable metadata (author, keywords, creation date) that you want to preserve during OCR. |
| Only OCR | Processes only pages that require OCR recognition. Skips pages that already have text layers without further analysis. | Large mixed documents where you want to process only the pages that need it, minimizing server load and processing time. |
| Skip OCR | Passes the document through the pipeline without performing any OCR recognition. Used for metadata manipulation only. | When you need to process a PDF for metadata or structural changes without altering the text content. Rarely used for OCR tasks. |
Default: Force OCR
If you are unsure which OCR type to choose, start with Force OCR. It is the most thorough option — it processes every page regardless of existing content. This ensures no page is missed and produces the most consistent results across the entire document.
Mixed Documents
For documents with a mix of scanned and digitally-born pages, use Skip All Text. This identifies pages with existing text layers and skips them, processing only the image-only pages. This is faster and avoids potential quality degradation on already-searchable pages.
Render & Output Types
Understanding hOCR rendering modes and PDF output formats for different use cases.
hOCR (Standard)
The default render type. Tesseract generates hOCR output — an XML format that stores recognized text with precise bounding box coordinates, font metadata, and confidence scores. Stirling PDF then merges this with the original PDF to create a searchable document. This produces clean, well-aligned text layers with consistent formatting across pages.
- Best for: General-purpose OCR, everyday documents
- Output quality: Clean text alignment, consistent spacing
- Processing speed: Standard
hOCR Original
A variant that preserves the original page layout more closely during the hOCR-to-PDF merge process. This mode prioritizes layout fidelity over text layer optimization. Use this when the standard hOCR rendering introduces spacing changes, font substitutions, or layout shifts that you want to avoid.
- Best for: Documents with complex layouts, tables, or precise formatting
- Output quality: Closer to original visual appearance
- Processing speed: Slightly slower due to layout preservation
PDF (Standard)
The default output format. Produces a standard PDF document compatible with all PDF viewers, browsers, and operating systems. This is the most universally compatible format and works with Adobe Acrobat, Preview, Chrome, Edge, Firefox, and every other PDF reader.
- Best for: Everyday use, sharing with others, general archiving
- Compatibility: 100% — works everywhere
- File size: Standard
PDF/A (Archival)
PDF/A is an ISO-standardized version of PDF designed for long-term digital preservation. It embeds all fonts,禁止s external dependencies, and uses only colorspace definitions that will render consistently over time. Required by many legal, government, and archival institutions for permanent document storage.
- Best for: Legal compliance, government submissions, archival, court filings
- Compliance: ISO 19005 standard
- File size: Slightly larger due to embedded fonts
Processing Options
Four toggle options that fine-tune the OCR pipeline for different document conditions.
Auto-Deskew Pages (Default: ON)
Automatically detects and corrects page rotation and skew before OCR processing. Scanned documents are frequently tilted by 1-5 degrees due to paper feed misalignment. Even a slight skew can cause OCR to misread character boundaries, dropping accuracy from 98% to 85%. Keep this enabled for all documents — it has virtually no performance cost and consistently improves results.
Clean Image Before OCR
Applies image preprocessing filters before Tesseract analyzes the page. This includes noise removal, contrast enhancement, and binarization. Useful for documents scanned at low resolution, photographed with a phone camera, or from old/damaged source material. Can slightly increase processing time but often improves accuracy on poor-quality scans by 5-15%.
Clean Image After OCR
Post-processes the output document to clean up visual artifacts introduced during the OCR recognition and PDF merge process. This can remove stray marks, correct minor alignment issues, and smooth out the text layer rendering. Generally optional — enable it if the output PDF shows visual artifacts or if the text layer appears misaligned with the original image.
Remove Images After Processing
Strips all embedded images from the output PDF, leaving only the text layer and basic formatting. This dramatically reduces file size — a 50MB scanned PDF can become 2-5MB after image removal. However, the visual content is lost. Use this when you need the text for search and indexing but do not need the original images (e.g., for text mining or data extraction pipelines).
The "Remove images after processing" option is irreversible. Once images are stripped from the output PDF, they cannot be recovered. Always keep a copy of the original scanned PDF before enabling this option. If you need both searchable text and the original images, process with image removal OFF and store both the original and the OCR output.
Tips for Better OCR Results
Practical techniques to maximize OCR accuracy and minimize errors.
Scan at 300 DPI or Higher
Resolution is the single most important factor for OCR accuracy. At 300 DPI, characters are large enough for Tesseract to analyze their shapes precisely. Below 200 DPI, accuracy drops significantly — especially for small fonts and fine details. For documents with small text (10pt or below), scan at 400-600 DPI for best results.
Ensure Good Lighting and Contrast
Even lighting across the page eliminates shadows and hotspots that confuse the OCR engine. For phone scans, use natural daylight or a desk lamp positioned to avoid reflections. High contrast between text and background is essential — black text on white paper is ideal. Colored backgrounds, gradients, or busy patterns reduce accuracy.
Keep the Document Flat
Curved, wrinkled, or folded pages cause character distortion. Press the document firmly against the scanner glass, or use a book scanner for bound documents. For phone photography, use a document scanning app that flattens perspective automatically. Wrinkles and folds create shadows that the OCR engine may interpret as character strokes.
Specify the Correct Language
Always enter the correct language code. Tesseract uses language-specific training data — using the wrong language causes the engine to match characters against incorrect patterns. For example, OCRing English text with the French model may confuse characters like "l" and "1" or "O" and "0". For multilingual documents, specify all applicable languages with plus signs.
Enable Auto-Deskew
Always keep auto-deskew enabled. Even a 2-3 degree tilt can cause Tesseract to misread character boundaries, leading to errors like "rn" being read as "m" or "cl" as "d". Deskew correction straightens the page before analysis, ensuring text lines are horizontal and characters are properly segmented.
Use Clean Before OCR for Poor Scans
If your source document is noisy, blurry, or low-quality, enable the "Clean image before OCR" option. This applies noise removal, contrast enhancement, and binarization before Tesseract analyzes the page. It can improve accuracy by 5-15% on documents with visible artifacts, speckles, or uneven lighting.
The best OCR results come from high-quality source documents. If possible, always scan at 300+ DPI with good lighting and flat pages. However, even imperfect scans can produce excellent results with the right settings — enable auto-deskew, specify the correct language, and use the clean-before-OCR option for noisy documents. For critical documents, process a test page first and verify accuracy before running the full batch.
Common Use Cases
Real-world scenarios where OCR transforms unsearchable scans into actionable documents.
Legal Documents
Court filings, contracts, depositions, and exhibits are frequently scanned and filed as image-only PDFs. OCR makes these documents searchable during legal discovery, enabling attorneys to find relevant passages, quotes, and evidence instantly. Courts increasingly require electronically searchable documents for e-discovery compliance.
Academic Research
University libraries have digitized millions of research papers, theses, and journals as scanned PDFs. OCR makes these archives searchable, enabling researchers to find specific studies, quotes, and data points across thousands of pages. Students can copy-paste citations instead of retyping them manually.
Medical Records
Hospitals and clinics scan patient records, lab results, and prescriptions into image-based PDFs. OCR enables healthcare providers to search for diagnoses, medications, and test results across patient histories. This is critical for emergency situations where quick information access saves lives.
Financial Documents
Invoices, receipts, bank statements, and tax documents are often received as scanned images. OCR extracts the text for data entry, accounting software import, and financial analysis. Accountants can search for specific transactions, amounts, and dates across months of scanned statements.
Historical Archives
Museums, libraries, and government agencies digitize historical documents, manuscripts, and records as scanned images. OCR makes these archives accessible to researchers and the public, enabling full-text search across centuries of recorded history. Combined with proper language codes, even old typefaces can be recognized.
Office Document Management
Offices generate mountains of paper that gets scanned into PDFs — memos, reports, meeting notes, correspondence. OCR transforms these into searchable digital assets that can be indexed by document management systems, found via desktop search, and referenced without opening each file individually.
Troubleshooting
Common problems when running OCR on scanned PDFs and how to fix them quickly.
Low OCR Accuracy or Garbled Text
Cause: Wrong language specified, low scan quality, or document contains unusual fonts.
Fix: Verify the language code matches the document language. For multi-language documents, specify all applicable languages with plus signs (e.g., eng+fra). If the scan is below 200 DPI, re-scan at 300+ DPI. Enable "Clean image before OCR" for noisy scans. For unusual fonts, Tesseract may struggle — try processing with the closest standard language model.
Upload Fails or File Rejected
Cause: File exceeds the 50MB size limit, is not a valid PDF, or the file is corrupted.
Fix: Check file size — compress the PDF using AFFLIGO's compress tool if needed. Verify the file has a .pdf extension and opens correctly in a PDF viewer. If the file is corrupted, try downloading it again from the source. Password-protected PDFs must be unlocked first before OCR processing.
Server Error or Processing Timeout
Cause: Server is under heavy load, file is very large, or network connection was interrupted.
Fix: Wait 30 seconds and try again. Large documents (50+ pages) may take longer to process — be patient during the processing phase. If the error persists, try a smaller file or split the document into smaller sections using AFFLIGO's split tool. Check your internet connection and try again.
Output PDF Is Much Larger Than Input
Cause: The OCR process added a text layer with high-resolution character data, increasing the file size.
Fix: This is normal — searchable PDFs are slightly larger than image-only PDFs. If the size increase is excessive (more than 2x), try processing with "Remove images after processing" to strip the original images and keep only the text layer. For standard documents, the size increase is typically 10-30%.
Text Layer Is Misaligned With Image
Cause: Auto-deskew incorrectly corrected the page rotation, or the document has mixed portrait/landscape pages.
Fix: Reprocess with auto-deskew disabled. For mixed-orientation documents, split portrait and landscape pages into separate files, process each separately, then merge the results. The hOCR Original render type may also produce better alignment for complex layouts.
Copy-Paste Still Does Not Work After OCR
Cause: The PDF viewer has not loaded the text layer, or the OCR output was not properly embedded.
Fix: Try opening the PDF in a different viewer (Adobe Acrobat, Chrome, Edge). Some viewers require you to explicitly enable text selection. If the issue persists, reprocess with Force OCR type to ensure a fresh text layer is created. Verify the output file size is larger than the input — a file with the same size likely did not receive the text layer.
OCR PDF: Frequently Asked Questions
Answers to the most common questions about OCR for scanned PDFs — covering tools, accuracy, privacy, and best practices.
A scanned PDF is an image of a document — it looks like text but contains no machine-readable characters. You cannot search, copy, or highlight text. A searchable PDF has a hidden text layer underneath the visual image, created by OCR. It looks identical but supports Ctrl+F search, copy-paste, screen readers, and digital indexing. OCR transforms the first into the second without changing the visual appearance.
Yes. AFFLIGO's OCR tool uses server-side processing via Stirling PDF. Your file is uploaded over an encrypted HTTPS connection, processed through the Tesseract OCR engine, and the result is returned to your browser. The server handles the computation so you do not need powerful hardware. Files are processed securely and are not stored after processing completes.
Accuracy depends on scan quality, language, and document complexity. A clean 300 DPI scan with standard fonts typically achieves 95-99% accuracy. Low-resolution scans, unusual fonts, handwritten text, or degraded documents may drop to 80-90%. Using the correct language code, enabling auto-deskew, and cleaning noisy scans before OCR can significantly improve results. For critical documents, always verify accuracy after processing.
Standard Tesseract OCR is designed for printed text and has limited capability with handwriting. Neat, consistent handwriting may be partially recognized, but cursive, irregular, or stylized handwriting will produce poor results. For handwritten documents, specialized handwriting recognition engines (like Google's Cloud Vision or Amazon Textract) perform better than standard Tesseract OCR.
No. Password-protected and encrypted PDFs must be unlocked before OCR processing. The server cannot open or modify a locked document. Use AFFLIGO's Unlock PDF tool to remove the password protection first, then process the unlocked file through OCR. After processing, you can re-apply password protection if needed.
AFFLIGO's OCR tool supports 38 languages including English, Spanish, French, German, Chinese, Japanese, Korean, Arabic, Portuguese, Italian, Dutch, Russian, Polish, Turkish, Greek, Thai, Vietnamese, Swedish, Norwegian, Danish, Finnish, Czech, Romanian, Hungarian, Indonesian, Malay, Croatian, Slovak, Slovenian, Catalan, Hebrew, Ukrainian, Bulgarian, Serbian, Lithuanian, Latvian, and Estonian. You can specify multiple languages for multilingual documents.
No. OCR adds an invisible text layer underneath the original image. The visual appearance is preserved exactly — the PDF looks identical before and after processing. The only change is the addition of a hidden text layer that enables search and copy-paste. The original image data is not modified. If you enable "Remove images after processing," the visual content will be stripped, but this is optional and off by default.
The maximum file size is 50MB per file. If your scanned PDF exceeds this limit, compress it using AFFLIGO's Compress PDF tool before processing. For very large documents (100+ pages), consider splitting the file into smaller sections using the Split PDF tool, processing each section separately, then merging the results. This also prevents timeout issues on slower connections.