Basics
Requests and pagination
Base URL, versioning, request formats, IDs and paging through lists.
Base URL
https://api.rowland.aiEvery endpoint is served over HTTPS.
Versioning
The version is part of the path: /v1. Within a version, Rowland may add endpoints, optional parameters and response fields, so ignore fields you do not recognize. Anything that would break existing code comes in a new version, announced ahead of time.
Requests
- File uploads are
multipart/form-data. Fields that hold a list or an object, such astagsormetadata, are JSON text. - Everything else takes and returns
application/json.
Objects and IDs
Every object says what it is in object, and its ID says it too:
| Prefix | Object |
|---|---|
doc_ |
A document |
ext_ |
An extraction of your own fields |
rev_ |
A revenue statement analysis |
cnv_ |
A chat conversation |
srh_ |
A search |
msn_ |
A WORK mission |
Times are ISO 8601 in UTC, such as 2026-11-20T15:04:05Z.
Pagination
Lists return a page at a time, newest first:
{
"object": "list",
"data": [],
"has_more": true
}Ask for up to 100 with limit (25 by default). When has_more is true, pass the last item's id as starting_after to get the next page.
Long-running work
Reading a document, extracting fields, analyzing a statement and running a mission all continue after the request returns 202 Accepted. Check the object's status, or use a webhook.
