Thai OCR — Extract Thai Text from Images and PDFs

Why Thai OCR Is Challenging

• Thai script has no spaces between words — word boundary detection requires linguistic analysis
• Tone marks and vowel signs stack above and below consonants in multiple vertical layers
• Vowels can appear before, after, above, or below the consonant they modify
• Distinguishing similar-looking Thai characters like ก/ถ, บ/ป, พ/ฟ in small font sizes
• Processing Thai text with embedded English words and Arabic numerals without script confusion
• Handling Thai numerals (๐-๙) alongside Western digits in the same document

How to Extract Thai Text (Step by Step)

1. Go to fastocr.org
2. Upload your Thai image or PDF. Language is detected automatically.
3. Wait for processing — images take seconds, PDFs show a progress bar.
4. Download results: searchable PDF, raw text file, or copy text directly.

Tips for Better Thai OCR Accuracy

1. Use 300+ DPI scans to preserve stacked tone marks and vowel signs above/below Thai consonants
2. Verify word segmentation in OCR output — Thai has no spaces so word boundaries may be incorrect
3. Check for correct positioning of vowels that appear before their consonant (e.g., เ, แ, โ)
4. For mixed Thai-English documents, review script boundaries to ensure clean transitions
5. Use high-contrast scans to preserve the fine details of stacked diacritical marks

Common Use Cases for Thai OCR

• Digitizing Thai legal documents, contracts, and court rulings
• Extracting text from Thai government forms and official certificates
• Converting scanned Thai academic papers and university publications
• Processing Thai business invoices, tax documents, and financial records
• Archiving Thai Buddhist texts, royal decrees, and historical manuscripts

Frequently Asked Questions