Developer — ClicheFactory

Getting Started

ClicheFactory extracts structured data from documents using AI. Send a PDF, image, or office file with a JSON schema — get clean, validated JSON back.

Choose Your Path

Python SDK

pip install clichefactory

For scripts, pipelines, and applications. Pydantic models, batch operations, trained extraction, and full programmatic control.

CLI

clichefactory extract

For terminal workflows, shell scripts, and CI/CD. Extract, convert, and diagnose from the command line.

MCP Server

Cursor · Claude Desktop

For AI coding assistants. Configure once and let your assistant extract data from documents as part of your dev workflow.

REST API

POST /v1/extract

For any language — curl, Postman, or your own HTTP client. Single-call extraction, no SDK required.

OpenClaw

Agent skill for OpenClaw-based automation.

Training

Build domain-specific extraction models. Currently BYOK-only.

Core Concepts

Modes, processing, schemas, file types, BYOK.

Quick Start

Install the package and run your first extraction. Pick your preferred tool:

            # Install

            pip install clichefactory

            clichefactory configure

            # Extract

            from clichefactory import factory

            client = factory(api_key="cliche-...")

            cliche = client.cliche({"type": "object", "properties": {"total": {"type": "number"}}})

            result = cliche.extract(file="invoice.pdf")

            print(result.total)

            # Install and configure

            pip install clichefactory

            clichefactory configure

            # Extract

            clichefactory extract invoice.pdf --schema schema.json

            curl -X POST "https://api.clichefactory.com/v1/extract" \

              -H "X-API-KEY: cliche-..." \

              -F "file=@invoice.pdf" \

              -F 'schema={"type":"object","properties":{"total":{"type":"number"}}}'