Edit | Fill PDF Forms and Modify DOCX with Natural Language | Reducto

Reducto leads independent benchmark on structured extraction with Deep Extract

Helping everyone from startups to Fortune 10 enterprises unlock their data.

Clients

Endpoints

/parse

Use when: Structured content from any document is needed for LLM or RAG use.
Output: Structured chunks with typed blocks, bounding boxes, and confidence scores.
How it works with Edit: Reads the document. Parse is the upstream step before Edit writes.

/extract

Use when: The fields to pull are defined and typed JSON is needed.
Output: Schema-typed JSON with optional citations on every value.
How it works with Edit: Returns data as JSON. Extract feeds values into Edit.

/split

Use when: One file contains multiple logical documents or sections.
Output: Page ranges for each section, with confidence scores.
How it works with Edit: Finds sections. Split page ranges can scope Edit calls.

/classify

Use when: Files need to be routed by type before processing.
Output: Best-matching category with per-criterion confidence.
How it works with Edit: Identifies document type. Classify picks which Edit recipe to use.

/edit

Use when: A PDF form needs filling or a DOCX needs updating.
Output: A downloadable edited file, plus a reusable form schema.

Instructions for Edits

  1. Natural-language instructions, not field IDs: Write what to fill, not where. “The applicant's full name” still finds the right field when the form is revised or the ID is opaque.
  2. Vision-based field detection: No embedded AcroForm required. Scans and widget-less PDFs still fill correctly.
  3. Reusable schemas for batches: Pass a saved form_schema to skip detection: ~3× faster and more consistent across thousands of the same form.
  4. Highlights changes in DOCX: Modified DOCX sections get highlighted so reviewers see exactly what changed.
  5. Overflow pages for long values: enable_overflow_pages: true appends an extra page when a value is longer than the field. Nothing gets truncated silently.
  6. Composable with the rest of the platform: Parse reads the document, Extract pulls values, Edit writes them back in.

Built for production