Processing Documents

Uploading via browser

Go to your parser's Document Inbox and drag files onto the upload area, or click to open the file picker. You can upload multiple files at once. Supported formats: PDF, PNG, JPG, WebP, DOCX, XLSX, PPTX, TXT, CSV, HTML, XML, JSON (max 10 MB).

Uploading via API

Use the POST /v1/invoices endpoint to upload documents programmatically, authenticated with an API key. See the Quickstart guide for full details.

Email import

Each parser can have its own private email address. Send or forward an email with PDF or image attachments to it and each document is parsed automatically, just like an upload. Only allowed senders can import. Set it up under Parser → Integrations → Import via Email; see the Integrations guide.

Batch uploads

Select multiple documents in the inbox and use Export Selected to export them together. For programmatic batch uploads, use POST /v1/invoices/batch (up to 10 files per request).

Processing status

Each document in the inbox shows its current status:

• Pending — queued for processing
• Processing — being processed now
• Complete — data extracted successfully
• Error — processing failed; click to see the reason

Re-processing a document

If a document didn't extract cleanly — a low-confidence field, a scan that came through blurry, or you've since fixed something on your end — you can re-run extraction on the original file without uploading it again:

• From the Document Inbox, open a document's ⋯ menu and choose Re-process document, or select several and use Bulk Actions → Reprocess.
• From the full-screen document viewer, click Reprocess in the toolbar.

Re-processing replaces the extracted fields and confidence scores with a fresh result — it doesn't create a new document. The status badge stays Processed either way, but a small Reprocessed tag appears next to it (in both the inbox and the document viewer) so you can tell it apart from a document's original parse. Hover the tag to see how many times it's been re-processed and when it last happened. Sample documents can't be re-processed since there's no original file behind them.

Pausing a parser

Go to a parser's Parser Settings → Danger Zone and click Pause Parser to temporarily stop it from accepting new documents — useful while you're reconfiguring fields or investigating a data-quality issue. While paused, both dashboard uploads and API requests for that document type are rejected with a clear message until you click Resume Parser. Documents already in the inbox, and re-processing them, are unaffected.