PerceptionDocs

Format support

Which file types each document operation reads.

An operation refuses a file type it doesn't read with unsupported_file_type. The document service decides a file's type from its first bytes, and the extension when those don't settle it.

File typeExtensionsParseSplitClassifyExtractGround
PDF.pdfYesYesYesYesYes
Spreadsheet.xlsx .xlsm .xls .csvYesYesYesYesYes
Parquet.parquetYesNoYesNoYes
Word document.docx .doc .docmYesYesYesYesYes
Image.png .jpg .jpeg .tiff .tif .webpYesYesYesYesYes
Bitmap.bmpNoYesYesYesYes
Presentation.pptx .pptYesYesYesYesYes
HTML.html .htmYesYesYesYesYes
Plain text.txt .log .xmlYesYesYesYesYes
Diagram.vsdx .vsdYesYesYesYesYes
Markdown.md .mdxYesYesYesYesYes
JSON.jsonYesYesYesYesYes
JSON Lines.jsonlNoNoNoNoYes
Email.eml .msg .oftYesYesYesYesYes
Video.mp4 .mov .avi .mkvYesNoNoNoYes
Audio.mp3 .wav .m4a .flac .mpga .oga .oggYesNoNoNoYes
  • Audio and video: Parse transcribes them. See Parse.
  • Markdown is passed through as it is, without parsing.
  • Large plain text: Parse reads .txt files up to 1 GiB and returns the complete text through document.content_url. Split, Extract and Ground read .txt files up to 256 MiB.