Documentation Index
Fetch the complete documentation index at: https://mintlify.com/reductoai/reducto-python-sdk/llms.txt
Use this file to discover all available pages before exploring further.
The Split API intelligently divides documents into sections based on content categories, making it easy to organize and process different parts of a document separately.
Basic Usage
from reducto import Reducto
client = Reducto()
response = client.split.run(
input="https://example.com/document.pdf",
split_description=[
{"name": "summary", "description": "Executive summary section"},
{"name": "financials", "description": "Financial data and tables"},
{"name": "risks", "description": "Risk factors and disclosures"}
]
)
print(response)
import asyncio
from reducto import AsyncReducto
client = AsyncReducto()
async def main():
response = await client.split.run(
input="https://example.com/document.pdf",
split_description=[
{"name": "summary", "description": "Executive summary section"},
{"name": "financials", "description": "Financial data and tables"},
{"name": "risks", "description": "Risk factors and disclosures"}
]
)
print(response)
asyncio.run(main())
Method Signature
client.split.run(
input: str,
split_description: list[SplitCategory],
parsing: ParseOptions | None = None,
settings: dict | None = None,
split_rules: str | None = None
) -> SplitResponse
Parameters
The URL of the document to split. You can provide:
- A publicly available URL
- A presigned S3 URL
- A
reducto:// prefixed URL from the /upload endpoint
- A
jobid:// prefixed URL from a previous parse invocation
- A list of URLs (for multi-document pipelines, V3 API only)
List of category definitions for splitting the document. Each category should have:
name: Category identifierdescription: Description of what content belongs in this category
Configuration options for parsing the document. If you’re passing in a jobid:// URL, this will be ignored.
Settings for split processing.
Natural language prompt describing rules for splitting the document.
Split Categories
Define categories based on document structure:
from reducto import Reducto
client = Reducto()
categories = [
{
"name": "introduction",
"description": "Opening remarks and overview"
},
{
"name": "methodology",
"description": "Research methods and approach"
},
{
"name": "results",
"description": "Findings, data, and analysis"
},
{
"name": "conclusion",
"description": "Summary and final thoughts"
}
]
response = client.split.run(
input="https://example.com/research-paper.pdf",
split_description=categories
)
# Access split sections
for section in response.sections:
print(f"{section.category}: {len(section.content)} chars")
import asyncio
from reducto import AsyncReducto
client = AsyncReducto()
async def main():
categories = [
{
"name": "introduction",
"description": "Opening remarks and overview"
},
{
"name": "methodology",
"description": "Research methods and approach"
},
{
"name": "results",
"description": "Findings, data, and analysis"
},
{
"name": "conclusion",
"description": "Summary and final thoughts"
}
]
response = await client.split.run(
input="https://example.com/research-paper.pdf",
split_description=categories
)
# Access split sections
for section in response.sections:
print(f"{section.category}: {len(section.content)} chars")
asyncio.run(main())
Custom Split Rules
Add natural language rules to guide the splitting process:
from reducto import Reducto
client = Reducto()
response = client.split.run(
input="https://example.com/contract.pdf",
split_description=[
{"name": "terms", "description": "Terms and conditions"},
{"name": "pricing", "description": "Pricing and payment terms"},
{"name": "warranties", "description": "Warranties and guarantees"}
],
split_rules="Split at major section boundaries. Keep related clauses together. Preserve the hierarchy of subsections."
)
Split with Parsing Options
Combine splitting with custom parsing configuration:
from reducto import Reducto
client = Reducto()
response = client.split.run(
input="https://example.com/document.pdf",
split_description=[
{"name": "text_sections", "description": "Text-heavy sections"},
{"name": "data_sections", "description": "Sections with tables and charts"}
],
parsing={
"enhance": {
"summarize_figures": True,
"agentic": ["table"]
},
"formatting": {
"table_output_format": "json",
"add_page_markers": True
}
}
)
Async Job Processing
For large documents, use async job processing:
from reducto import Reducto
client = Reducto()
# Start an async split job
job = client.split.run_job(
input="https://example.com/large-document.pdf",
split_description=[
{"name": "section1", "description": "First section"},
{"name": "section2", "description": "Second section"}
],
async_={
"webhook": {"url": "https://example.com/webhook"}
}
)
print(f"Job ID: {job.job_id}")
# Poll for results
result = client.job.get(job.job_id)
Financial Document Example
Split a financial report into meaningful sections:
from reducto import Reducto
client = Reducto()
categories = [
{
"name": "executive_summary",
"description": "High-level overview and key highlights"
},
{
"name": "financial_statements",
"description": "Income statement, balance sheet, cash flow"
},
{
"name": "md_and_a",
"description": "Management discussion and analysis"
},
{
"name": "footnotes",
"description": "Accounting notes and disclosures"
},
{
"name": "risk_factors",
"description": "Risk disclosures and forward-looking statements"
}
]
response = client.split.run(
input="https://example.com/annual-report.pdf",
split_description=categories,
split_rules="Preserve table integrity. Keep footnote references with their corresponding sections."
)
# Process each section separately
for section in response.sections:
print(f"\n=== {section.category.upper()} ===")
print(section.content[:200] + "...")
Reusing Parsed Documents
Split a document that was previously parsed:
from reducto import Reducto
client = Reducto()
# First parse the document
parse_response = client.parse.run(
input="https://example.com/document.pdf",
formatting={"add_page_markers": True}
)
# Then split using the job ID (no re-parsing needed)
split_response = client.split.run(
input=f"jobid://{parse_response.job_id}",
split_description=[
{"name": "part1", "description": "First part"},
{"name": "part2", "description": "Second part"}
]
)