# shcherbak-ai/contextgem

ContextGem: Effortless LLM extraction from documents

Repository: https://github.com/shcherbak-ai/contextgem
Canonical: https://ross.abutalabs.com/products/contextgem
Homepage: https://contextgem.dev/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: ai, contract-analysis, data-extraction, document-intelligence, generative-ai, legaltech, llm, llm-extraction, llm-framework, llm-pipeline, llms, nlp, prompt-engineering, text-analysis, unstructured-data, autoprompting
Last push: 2026-08-13T15:57:23+00:00

## Health v2 (maintenance only)
Score: 87/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 97, release rhythm 97, longevity 49
- inputs: {"age_days": 691, "days_push": 20, "days_rel": 20, "gap_med": 5.0, "n_releases_24m": 51}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1993, forks 183 (observed 2026-08-28T04:06:03.281536+00:00)

## What it is
ContextGem is an open-source Python LLM framework for extracting structured data and insights from documents with minimal code. It provides declarative APIs for defining aspects and concepts, reusable extraction pipelines, and support for cloud and local LLM providers.

## Use cases
- extract structured data from contracts with an LLM
- pull specific clauses out of legal agreements
- extract dates, numbers, and JSON objects from documents
- analyze long PDFs and reports for key sections
- build reusable document extraction pipelines
- parse unstructured text into structured fields without prompt engineering
- identify topics and sections in research papers or financial reports

## When to choose
- you need structured extraction (aspects, concepts, typed fields) from documents with minimal boilerplate
- you want automatic prompting and justification/reference mapping out of the box
- you work with legal, financial, or other long-form documents and need semantic segmentation rather than naive chunking
- you want a Python library that works with multiple cloud and local LLM providers

## When to avoid
- you need a general-purpose RAG system with vector search and retrieval over a corpus
- you need high-throughput bulk document OCR or image processing rather than LLM-based extraction
- you want a no-code UI or hosted service instead of a Python library
- your extraction tasks are simple regex or template parsing that doesn't need an LLM

## Facets
- artifact type: library
- maturity: active
- function: llm-inference, nlp, rag, prompt-engineering, serialization
- domain: large-language-models, artificial-intelligence, legal, data-science
- platform: python, cross-platform
- tags: document-intelligence, information-extraction, aspect-extraction, concept-extraction, contract-analysis, structured-data, llm-framework, legaltech, zero-shot-extraction, multimodal, natural-language-processing, documents

## Member repositories
- shcherbak-ai/contextgem (main) score 87

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:03.281536+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 2026-08-30T03:02:36.116967+00:00, confidence not recorded.
  - readme: https://github.com/shcherbak-ai/contextgem (fetched 2026-08-28T04:06:03.281536+00:00, sha 3dffe4b191cc)
  - homepage: https://contextgem.dev/ (fetched 2026-08-29T10:42:25.403073+00:00, sha 2aa5f69149b2)
  - site_page: https://contextgem.dev/installation (fetched 2026-08-29T10:42:25.412134+00:00, sha 3e7c5bd1fdab)
  - site_page: https://contextgem.dev/quickstart (fetched 2026-08-29T10:42:25.413921+00:00, sha 02390ad2f9e2)
  - site_page: https://contextgem.dev/optimizations/optimization_long_docs (fetched 2026-08-29T10:42:25.416202+00:00, sha 5177d9e46df8)
  - registry_pypi: https://pypi.org/pypi/contextgem/json (fetched 2026-08-29T10:42:25.417845+00:00, sha 0a31e4472ecf)
- Data as of 2026-08-30T08:39:29.467469+00:00.