Introducing s-1: Sievly’s new document reading model
Documentation
Sievly API documentation
The Sievly API turns documents into structured data. This page covers what you need for a first integration. Every endpoint takes a file or URL and returns JSON with citations.
Quickstart
Install the Python or Node SDK, add your API key as an environment variable and call Parse on a file. New workspaces start with free credits.
pip install sievly
export SIEVLY_API_KEY=”your-key”
from sievly import Sievly
result = Sievly().parse(“invoice.pdf”)
Authentication
Every request is authenticated with an API key sent as a bearer token. Create and rotate keys in your workspace settings. Keep keys on the server and never ship them in client code.
Authorization: Bearer $SIEVLY_API_KEY
Endpoints
Each endpoint accepts a file upload or a document URL. Responses include page numbers and bounding boxes for every block or value.
POST /parse
Structured content, chunks, tables and figures
POST /extract
Values for a schema you define
POST /split
Page ranges for each document in a file
POST /classify
A category and confidence for a document
POST /edit
An edited file from a plain-language instruction
Errors and limits
Errors return a JSON body with a code and a readable message. Rate limits depend on your plan; Growth and Enterprise include premium limits and priority requests. Long-running jobs can be sent asynchronously with a webhook for the result.