How do I upload large files through an API without timeouts?
Request a presigned upload URL from FormKiQ, then send file bytes directly to Amazon S3 using the returned URL and headers. This avoids carrying the file through the inline API request. Transfer limits, URL expiry, site configuration, and downstream processing limits still apply; processing readiness must be checked separately after upload.
Use a two-step transfer
Call POST /documents/upload with the path, content type, and intended siteId. Record the returned documentId, then upload to url. Apply every returned signed header exactly. Do not attach your FormKiQ JWT to the S3 request or change signed URL parameters.
The documented v1.19.2 flow is a single presigned S3 PUT. It is not a multipart-upload API. Check the release-specific limits reference rather than assuming all S3 upload techniques are exposed through FormKiQ.
Handle failures deliberately
An expired URL needs a new authorized request. A browser upload also needs the appropriate bucket CORS configuration; an API Gateway CORS setting alone does not configure the S3 transfer.
Distinguish an interrupted transfer from delayed processing. Reconcile document state before creating another document after an uncertain outcome. Treat a presigned URL as a temporary credential and keep it out of public logs, screenshots, and analytics. Test the largest expected file with the actual OCR, scanning, and search pipeline.
Edition and configuration
The upload URL flow is part of the document APIs. Configured site and processor limits can be lower than storage-service limits, regardless of edition.