Skip to content

Supported Document Formats

PrimeThink supports a wide variety of document formats for upload and processing. When you upload a document, the system automatically extracts the text content and makes it available to AI assistants in your chats.

Document Text Extraction Process

When you upload a document, PrimeThink automatically extracts text content in Markdown format for supported file types. This extraction process is asynchronous and happens in the background.

How Extraction Works

  1. Upload: When a file is uploaded, it receives an initial status of "Added"
  2. Processing: The system begins extracting text content asynchronously
  3. Completion: Once extraction is complete, the status changes (e.g., to "Processed", "Ready", or "Indexed")
  4. Error Handling: If extraction fails, the status becomes "Error"

Extraction Timing

  • Processing Time: Varies from a few seconds to longer depending on:
  • File size (larger files take longer)
  • File complexity (scanned PDFs with OCR take longer than text-based documents)

  • System load

  • Status Indicators:

  • "Added" = Extraction not yet started or in progress
  • "Error" = Extraction failed
  • Any other status ("Processed", "Ready", "Indexed", etc.) = Extraction complete, text available

Extracted Content Format

The extracted text is converted to Markdown format, which includes: - Document structure (headings, paragraphs) - Tables (for spreadsheets and documents with tables) - Basic formatting (bold, italic, lists) - Text from images (via OCR or AI vision) - Transcriptions (for audio/video files)

Checking Extraction Status

To check if a document's text has been extracted: - Check the document's status field in API responses - Use pt.getDocumentStatus(docId) in live apps - Any status other than "Added" or "Error" indicates extraction is complete

For programmatic workflows in live apps, see Data Management API for details on checking status and retrieving extracted text.

Document Files

Format Extensions Description
PDF .pdf Portable Document Format with OCR support for scanned documents
Word .doc, .docx Microsoft Word documents (both legacy and modern formats)
Plain Text .txt Simple text files
Markdown .md Markdown formatted text
HTML .html, .htm Web pages and HTML content

PDF Processing: PrimeThink can extract text from both digital PDFs and scanned documents using optical character recognition (OCR).

Spreadsheets

Format Extensions Description
Excel .xls, .xlsx Microsoft Excel spreadsheets (both legacy and modern formats)
CSV .csv Comma-separated values

Presentations

Format Extensions Description
PowerPoint .ppt, .pptx Microsoft PowerPoint presentations (both legacy and modern formats)

Data Files

Format Extensions Description
JSON .json JavaScript Object Notation
XML .xml Extensible Markup Language

Images

Format Extensions Description
JPEG .jpg, .jpeg Standard image format
PNG .png Portable Network Graphics
GIF .gif Graphics Interchange Format
BMP .bmp Bitmap images
WebP .webp Modern web image format

Image Processing: PrimeThink uses AI vision to extract text from images and generate descriptions of image content.

Audio and Video

Format Extensions Description
MP3 .mp3 Audio files
M4A .m4a Apple audio format
WAV .wav Waveform audio
MP4 .mp4 Video files

Audio/Video Processing: Audio and video files are automatically transcribed to text, making spoken content searchable and accessible to AI assistants.

Email Files

Format Extensions Description
Email .eml Standard email message files

Email Processing: The system extracts email headers (From, To, Subject, Date), body content, and lists any attachments.

Archives

Format Extensions Description
ZIP .zip Compressed archive files

Archive Processing: ZIP files are automatically extracted, and each file within the archive is processed individually.

Web Content

Type Description
URLs Paste any web URL to automatically fetch and extract its content
YouTube Videos YouTube links are processed to extract video metadata, descriptions, and transcripts when available

File Size Limits

  • Maximum file size: 50 MB per file
  • ZIP archives: Subject to limits on total uncompressed size and file count

Tips for Best Results

  1. Use text-based formats when possible - Digital documents (like .docx) typically yield better results than scanned PDFs.

  2. Ensure good image quality - For images containing text, higher resolution images produce more accurate text extraction.

  3. Check audio quality - Clear audio recordings with minimal background noise transcribe more accurately.

  4. Name files descriptively - Use meaningful filenames to help organize and identify your documents.