Build with the DocsFlow AI API

Convert PDFs, scans, invoices, IDs, and websites into clean JSON with one REST call. Covers quickstart, endpoints, authentication, and SDKs.

Jump to quickstartGet API key free
Jump to:QuickstartAPI ReferenceSDKsAuthentication

Extract your first invoice in 5 minutes

One REST call, clean JSON back. No templates, no training, no configuration required.

01

Create a free account

First 100 documents/month free — no credit card.

02

Copy your API key

Available in your dashboard under API Keys.

03

Send a document

POST to /v1/parse with your file or URL.

04

Receive structured JSON

Fields, tables, line items, and confidence scores.

import requests

resp = requests.post(
    "https://api.docsflowai.com/v1/parse",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    files={"document": open("invoice.pdf", "rb")},
    json={"output": "json"},
)
data = resp.json()
print(data["fields"])
# {"invoice_number": "INV-2026-001", "vendor": "Acme Corp",
#  "amount": 4250.00, "currency": "USD", "due_date": "2026-04-15"}

Five endpoints, unlimited possibilities

A minimal, predictable surface. Every response follows the same shape — no surprises in production.

Secure By Default

API keys are scoped, rotatable, and never leave your control. All processing happens in isolated containers — your documents are never stored.

Bearer token in header

Pass your API key as Authorization: Bearer YOUR_API_KEY in every request. Never in query parameters.

TLS required

All API traffic must use HTTPS/TLS 1.2+. Plain HTTP requests are rejected with 400.

Key rotation

Rotate keys from the dashboard without downtime. Old keys stay valid for 24 h after rotation.

Zero data retention

Documents are processed in isolated containers and deleted immediately after the response is returned.

HTTP headers
POST https://api.docsflowai.com/v1/parse

Authorization: Bearer sk_live_••••••••••••••••
Content-Type: multipart/form-data
X-Request-ID: req_7mNqR4wX

Auth error codes

401Missing or malformed API key
403Key revoked or plan limit reached
429Rate limit exceeded — retry after header

Integrate In Minutes, Not Days

Official SDKs, a CLI, webhooks, and an OpenAPI spec — built for developers who value predictability.

Official SDKs

PY
PythonWorks with Python 3.8+
pip install docsflow
NO
Node.jsESM + CommonJS
npm install @docsflow/sdk
GO
GoGo 1.20+
go get github.com/docsflow/go
RU
RubyRuby 3.0+
gem install docsflow

Full API reference is available as machine-readable llms.txt for AI tools, agents, and IDE copilots.

DocsFlow REST API
POST/v1/ocr/extract
{
  "file_url": "https://cdn.acme.com/inv-2847.pdf",
  "document_type": "invoice",
  "output_schema": "standard"
}
Ready to send
Response
CLI tool
Extract documents directly from your terminal. Pipe output to jq, csvkit, or any shell tool.
OpenAPI 3.0 spec
Full machine-readable spec. Import into Postman, Insomnia, or generate your own typed client.
Webhooks
Push extraction results to any URL the moment a job completes. Retry logic built in.
Streaming API
Stream large batch results as NDJSON. No need to wait for the full job to finish.
DocsFlow AI API Guide: Endpoints, Auth & SDKs
The DocsFlow AI API converts PDFs, scans, invoices, IDs, and websites into structured JSON through five core endpoints, Bearer-token authentication, and official SDKs for Python, Node.js, Go, and Ruby. This guide explains the endpoints, how authentication works, what each response contains, and how the API protects document data.
What is the DocsFlow AI API?
The DocsFlow AI API is a REST interface that converts PDFs, scans, invoices, IDs, contracts, and websites into structured JSON with a single HTTP call. It pairs OCR with AI layout understanding so responses contain validated fields, tables, line items, and per-field confidence scores — not raw text. Developers call https://api.docsflowai.com with a Bearer API key and receive machine-readable output ready for databases, ERPs, and automation workflows.
How do I get started with the DocsFlow AI API?
Create a free account to generate a scoped API key, then send your first request to POST /v1/parse with a document file or URL. You receive a job ID immediately and a structured JSON response once processing completes. The free Starter plan includes 100 documents per month, and official SDKs for Python, Node.js, Go, and Ruby remove most setup work.
Which API endpoints does DocsFlow AI provide?
The API exposes five core endpoints. POST /v1/parse extracts structured data from documents. POST /v1/crawl starts an async website crawl, including JavaScript-rendered pages. POST /v1/parse/url parses a document directly from a download URL without uploading it. GET /v1/jobs/:id polls asynchronous job status and results, and GET /v1/account returns plan limits and usage.
How does API authentication work?
Pass your API key as an Authorization: Bearer header on every request — never in query parameters. All traffic must use HTTPS with TLS 1.2 or higher. Keys are scoped to read-only, write, or admin permissions, can be rotated from the dashboard without downtime, and are rate-limited per key. Revoked or over-limit keys return 403, and rate limits return 429 with a retry-after header.
How does the API handle security and data privacy?
Every document is processed in an isolated container and deleted immediately after the response is returned — DocsFlow AI keeps zero data retention. Requests require TLS 1.2+, API keys are scoped and rotatable, and processing runs under the same SOC 2 Type II, GDPR, and HIPAA-compatible controls as the rest of the platform. No third parties access your documents at any stage.
DocsFlow AI API at a glance
Core endpoints
5
Supported formats
40+
Official SDKs
Python · Node · Go · Ruby
Authentication
Bearer token
Response formats
JSON · CSV · XML
Data retention
Zero
 FAQ
Documentation & API questions answered
What developers ask before building with the DocsFlow AI API.
What is the DocsFlow AI API?
The DocsFlow AI API is a REST interface that converts PDFs, scans, invoices, IDs, contracts, and websites into structured JSON with a single HTTP call. Responses include validated fields, tables, line items, and per-field confidence scores, ready for databases, ERPs, and automation workflows.
How do I get an API key?
Create a free account at https://alpha.docsflowai.com/login, open Settings, and generate a scoped API key. The Starter plan includes 100 documents per month at no cost and no credit card is required. Keys are scoped to read-only, write, or admin permissions.
Which endpoints does the DocsFlow AI API provide?
Five core endpoints: POST /v1/parse extracts structured data from documents, POST /v1/crawl starts an async website crawl, POST /v1/parse/url parses a document from a download URL, GET /v1/jobs/:id polls job status, and GET /v1/account returns plan limits and usage.
How do I authenticate API requests?
Pass your API key as an Authorization: Bearer header on every request, never in query parameters. All traffic must use HTTPS with TLS 1.2 or higher. Keys are scoped, rotatable from the dashboard without downtime, and rate-limited per key.
Which SDKs are available for DocsFlow AI?
Official SDKs cover Python (3.8+), Node.js (ESM and CommonJS), Go (1.20+), and Ruby (3.0+). A CLI, a machine-readable OpenAPI 3.0 spec, webhooks with built-in retry, and a streaming NDJSON endpoint are also available.
What document formats does the DocsFlow AI API support?
The API accepts over 40 formats, including PDF (digital and scanned), PNG, JPG, TIFF, WEBP, DOCX, XLSX, and PPT. Document layouts are auto-detected, so no templates are required for invoices, contracts, IDs, and medical records.
Does the DocsFlow AI API store my documents?
No. Every document is processed in an isolated container and deleted immediately after the response is returned. DocsFlow AI keeps zero data retention — no backups, caching, or content logs of your documents at any stage.
What response formats does the API return?
Responses default to JSON and can be requested as CSV or XML. Each extraction includes labelled fields, table rows, line items, confidence scores, source page references, and bounding boxes. Custom extraction schemas define the exact fields and types returned.
Ready to start extracting?
First 100 documents per month are free. No credit card required. Get your API key in under a minute.
Get API key free View pricing
WhatsApp