# Ontos-AI/knowhere

Knowhere extracts, parses, and outputs structured chunks ready for AI Agents and RAG.

Repository: https://github.com/Ontos-AI/knowhere
Canonical: https://ross.abutalabs.com/products/knowhere
Homepage: https://knowhereto.ai
Language: Python
License: Apache-2.0
License Family: permissive
Topics: ai-agents, claude-code, cursor, langchain, rag-pipeline, skills, agent-harness, codex, document-intelligence, mutli-agent, vector-database
Last push: 2026-08-26T02:54:17+00:00

## Health v2 (maintenance only)
Score: 81/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 99, longevity 8
- inputs: {"age_days": 125, "days_push": 7, "days_rel": 8, "gap_med": 0, "n_releases_24m": 38}
- flags: young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2673, forks 292 (observed 2026-08-28T04:07:09.279664+00:00)

## What it is
Knowhere is a document ingestion and parsing platform that transforms unstructured documents (PDF, DOCX, XLSX, PPTX, images) into structured, RAG-ready chunks with preserved hierarchy, tables, and metadata. It is offered as a managed REST API (Chunking-as-a-Service), an open-source self-hosted stack, and an MCP server for AI agents like Cursor, Claude Code, and Codex.

## Use cases
- parse resumes from pdfs into structured chunks
- prepare messy PDFs for RAG pipelines
- extract tables and layouts from documents for LLMs
- chunk long technical documents with hierarchy preserved
- give AI agents document parsing via MCP
- build agentic RAG over unstructured documents
- convert docx and pptx files into clean JSON chunks

## When to choose
- you need high-quality, structure-preserving chunking for RAG or agents
- you want an MCP server so coding agents can parse and search documents
- you deal with complex PDFs with tables, merged cells, or hundreds of pages
- you want a managed API instead of maintaining your own parsing pipeline

## When to avoid
- you only need simple plain-text extraction from clean documents
- you need fully offline processing and cannot use a cloud API or self-host
- your documents are already well-structured markdown or JSON
- you need formats not yet supported like epub or html

## Facets
- artifact type: service
- maturity: active
- function: parser, rag, etl, pdf, ocr, mcp, sdk, search-engine
- domain: artificial-intelligence, pdf, developer-tools, apis
- platform: python, self-hosted, cli
- tags: chunking-as-a-service, document-parsing, agentic-rag, document-intelligence, vectorless-rag, hierarchical-chunking, mcp-server, rest-api, retrieval-augmented-generation, ai-agents, natural-language-processing, nodejs, docker, web-server

## Member repositories
- Ontos-AI/knowhere (main) score 81

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:09.279664+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:18:04.932457+00:00, confidence not recorded.
  - readme: https://github.com/Ontos-AI/knowhere (fetched 2026-08-28T04:07:09.279664+00:00, sha 4f2196b9750c)
  - homepage: https://knowhereto.ai (fetched 2026-08-29T10:00:38.734124+00:00, sha ad2ad83070db)
  - site_page: https://docs.knowhereto.ai/ (fetched 2026-08-29T10:00:38.736635+00:00, sha 2a2d730299c8)
  - site_page: https://docs.knowhereto.ai/mcp (fetched 2026-08-29T10:00:38.738292+00:00, sha b9b54943b812)
- Data as of 2026-08-30T08:39:29.467469+00:00.