Split API

Break document bundles into typed sub-documents

Boundary detection finds where each document starts and ends in a compound file, labels it with your types, and returns each one as a standalone file.

What it does

  • Typed page ranges

    Each split returns its type, start and end page, and a file ID you can pass straight to Extract or Classify.

  • Identifiers per document

    An identifier key pulls a value such as an invoice number from each split; strict mode merges adjacent splits that share it.

  • Performance or Light

    Choose the engine per splitter to trade accuracy against cost on long packets.

Under the hood

Segment multi-document files into individual subdocuments.

Split
  • Large document splitting

    High-precision splitting that maintains accuracy on 2,000+ page files.

  • Instance detection

    Detect unique identifiers to separate multiple instances of the same document type, like splitting 50 invoices by invoice number.

  • Intelligent boundary handling

    Smart overlap handles mid-page boundaries while keeping context intact.

  • Cost-optimized splitting

    Toggle cost-optimized splitting for bulk jobs, or high precision for tough files.

Benchmarks

  • PoliTax Split

    72.48%Best boundary-detection F1 with Extend's harness

    The 30 largest compound tax filings in the public PoliTax corpus.

    PoliTax Split report

  • PoliTax Split

    +8.3 to +28.4F1 points over direct use of the same model

    The harness catches about two thirds or more of boundaries for every underlying model.

    PoliTax Split report

How to get started

  1. Create an account

    Sign up at dashboard.extend.ai. New accounts get 10,000 free credits.

    Copy an API key from the Developers page and export it as EXTEND_API_KEY.

  2. Choose how to call Split

    Studio: try Split in the browser with no code.

    MCP: connect Claude, Cursor, or any MCP client to https://mcp.extend.ai/mcp.

    CLI: install with curl -fsSL https://extend.ai/install.sh | sh, then run extend setup.

    SDKs: pip install extend-ai or npm install extend-ai; Java and Go SDKs are also available.

  3. Process your first document

    Define the document types to look for. At least one type must be other.

    curl -X POST https://api.extend.ai/split \
      -H "Authorization: Bearer $EXTEND_API_KEY" \
      -H "x-extend-api-version: 2026-02-09" \
      -H "Content-Type: application/json" \
      -d '{ "file": { "url": "https://extend-public-files.s3.us-east-2.amazonaws.com/loan_application.pdf" }, "config": { "splitClassifications": [ { "id": "invoice", "type": "invoice", "description": "An invoice or bill for goods or services." }, { "id": "other", "type": "other", "description": "Any other document type." } ] } }'
  4. Move to production

    Send production traffic to POST /split_runs; the synchronous POST /split times out after 5 minutes.

Pricing

From 0.5 credits per page with Light Split, with parsing billed on top.

See full pricing

Use cases

  • Lab & genomic reports

    Find the sections of 50-page reports that hold the fields you need before extracting.

    Flatiron Health case study
  • Loan & mortgage packets

    Separate statements, pay stubs, and disclosures in one closing packet.

    Real estate
  • Claims bundles

    Break insurance, legal, and healthcare claim packets into typed documents.

    Healthcare
  • Compound tax filings

    Find form boundaries across hundreds of pages of schedules and supplements.

    PoliTax Split benchmark
  • Batch scans

    Turn one scanned stack into a file per document, ready to route.

    Splitter product spotlight
  • Freight paperwork

    Separate bills of lading, invoices, and customs forms sent as one PDF.

    Logistics
cta-background

( fig.11 )

Turn your documents into high quality data