AI · Web app
Reducto
Agentic document parsing and extraction platform for AI pipelines.
Reducto is a document-processing API designed to parse, split, extract, and edit data from complex documents like PDFs, spreadsheets, and scanned forms. It combines OCR with vision language models to preserve document structure, tables, and formatting for downstream LLM workflows. It also supports on-premise or VPC deployments for strict compliance needs.
What it's for
- Parse complex multi-column PDFs and financial tables into LLM-ready formats
- Extract structured JSON data from invoices and onboarding forms using defined schemas
- Split large combined document batches into individual files using layout heuristics
- Fill in blanks and form fields automatically without fixed templates
Features
- Multi-pass OCR combined with vision language models (r-1 model)
- Document splitting with layout-aware heuristics
- Schema-level structured data extraction
- Template-free automated form editing and filling
- Granular bounding-box citations for extracted data
- Multilingual support across 100+ languages including mixed text
- Python and Node.js SDK support
- On-premise / self-hosted VPC deployment option
Pricing
Freemium. Free trial tier available; pay-as-you-go standard tier starts around 1¢ per page; custom pricing for Enterprise.
Details
- Type
- Web app
- Category
- AI
- Topic
- Document Processing
- Open source
- No
- Platforms
- API, Web, Self-hosted
- For
- Developers and AI engineers building RAG pipelines and document automation workflows.
- Alternative to
- LlamaParse, AWS Textract, Unstructured, Google Document AI
- Tags
- ocr, document parsing, data extraction, llm, rag, pdf
Why it was saved
- @d__raptis: Real product shown in the video — an AI document-processing API worth knowing about.
Related
Open-source vision-language model for parsing multi-page documents in one pass
Practical architectural strategies to fix common failure modes in RAG systems
Private, on-device AI assistant to chat with documents and local models.
RAG framework using continuous latent reasoning and embedding-based context compression
Desktop app to run open-source language models locally and completely offline
Open-source persistent memory layer for autonomous AI agents using knowledge graphs
Updated