Configuration Overview

Every Reducto endpoint exposes configuration options that control how documents are processed. This section covers all available configurations, from parse-level OCR and layout settings through extraction schemas and workflow orchestration.

Configuration by Endpoint

Parse
Extract
Split
Classify
Edit

Parse converts documents into structured content. Options are grouped by purpose:

Group	Purpose	Pages
`enhance`	AI-powered accuracy	Agentic Modes, Chart Extraction
`retrieval`	RAG optimization	Chunking Methods
`formatting`	Detecting styling & output format	Table Formats, Additional Document Data
`spreadsheet`	Excel/CSV handling	Spreadsheet Processing
`settings`	Processing controls	Processing Settings, Page Ranges

result = client.parse.run(
    input=upload,
    enhance={...},
    retrieval={...},
    formatting={...},
    spreadsheet={...},
    settings={...}
)

Extract pulls structured data from documents using a JSON schema.

Group	Purpose	Pages
`instructions`	Schema and system prompt	(base config)
`settings`	Citations, array extraction	Array Extraction, Citations
`parsing`	Document processing	All Parse options

result = client.extract.run(
    input=upload,
    instructions={"schema": {...}, "system_prompt": "..."},
    settings={"array_extract": True, "citations": {"enabled": True}},
    parsing={...}
)

Split divides documents into logical sections.

Group	Purpose	Pages
`split_description`	Section definitions	Split Configuration
`split_rules`	Splitting logic prompt	Split Configuration
`settings`	Table handling	Split Configuration
`parsing`	Document processing	All Parse options

result = client.split.run(
    input=upload,
    split_description=[{"name": "...", "description": "..."}],
    split_rules="...",
    settings={"table_cutoff": "truncate"}
)

Classify determines document type based on natural language criteria.

Group	Purpose	Pages
`classification_schema`	Category definitions	Classify Configuration
`page_range`	Pages used as context	Classify Configuration

response = client.classify.run(
    input=upload,
    classification_schema=[
        {"category": "invoice", "criteria": ["billing info", "itemized charges"]},
        {"category": "contract", "criteria": ["legal terms", "signatures"]},
    ]
)

Edit fills forms and modifies documents.

Option	Purpose	Pages
`edit_instructions`	Natural language instructions	(base config)
`form_schema`	Pre-defined field locations	Form Schema
`edit_options`	Highlight color, overflow	(base config)

result = client.edit.run(
    document_url=upload,
    edit_instructions="Fill name: John Doe, date: 2024-01-15",
    form_schema=[...],
    edit_options={"color": "#FF0000"}
)

Common Patterns

RAG-optimized parsing

Variable chunking with embedding optimization for vector search:

result = client.parse.run(
    input=upload,
    retrieval={
        "chunking": {"chunk_mode": "variable", "chunk_size": 1000},
        "embedding_optimized": True
    },
    formatting={"table_output_format": "dynamic"}
)

High-accuracy processing

Enable agentic mode for both text and tables:

result = client.parse.run(
    input=upload,
    enhance={
        "agentic": [{"scope": "text"}, {"scope": "table"}]
    }
)

Complete extraction with citations

Array extraction with source locations for long documents:

result = client.extract.run(
    input=upload,
    instructions={"schema": schema},
    settings={
        "array_extract": True,
        "citations": {"enabled": True}
    }
)

Migrating from v2

If you’re using the legacy configuration format, use this converter to transform your v2 config to v3:

See the Migration Guide for complete mapping tables and examples.

​Configuration by Endpoint

​Common Patterns

​Migrating from v2

Configuration by Endpoint

Common Patterns

Migrating from v2