#1919th of 24 in Data and scraping for AIFor data and documents, documents and ocr
unstract
Extracts structured JSON from PDFs and scans using LLM prompts
Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 7.3k
- License
- AGPL-3.0
- Last commit
- Oct 2026
- Last release
- Oct 2026
- Language
- Python
Missed the latest GitHub check; these facts may be older.
Overview
Unstract extracts structured JSON from documents such as PDFs, images and scans. Extraction schemas are defined as natural-language prompts in Prompt Studio, then deployed as a REST API or an ETL pipeline that loads results into a warehouse. It also offers an MCP server and an n8n node, and works with OpenAI, Anthropic, Bedrock, Ollama and other LLM providers.
Who it is for: Teams extracting structured data from invoices, forms and scanned documents
Strengths
- Prompt Studio defines extraction schemas in natural language instead of per-vendor templates
- Deploys as a REST API or ETL pipeline from S3, GCS, Drive, SFTP to Snowflake, BigQuery, Postgres
- Supports many LLMs, including local Ollama, and vector DBs such as Qdrant, Weaviate, Milvus
- Ships an MCP server and an n8n custom node for agent and workflow integration
Weaknesses
- Heavy stack: Django, Celery, FastAPI, Redis, RabbitMQ and PostgreSQL; 8 GB RAM minimum
- LLMChallenge, SinglePass extraction, human review and SSO/RBAC are cloud or enterprise only
- README lists Linux or macOS only; no Windows setup is documented
- Default login unstract/unstract must be changed; losing ENCRYPTION_KEY locks adapter credentials
What it needs
- RAM ≥ 8 GB
- no GPU
- Needs Docker, Docker Compose, PostgreSQL, Redis, RabbitMQ, LLM provider, Vector database, Text extractor (LLMWhisperer, Unstructured.io or LlamaIndex Parse)
- Models: OpenAI, Azure OpenAI, OpenAI Compatible, Anthropic Claude, AWS Bedrock
Also in Data and scraping for AI
See all 24| Rank | Project | Score |
|---|---|---|
| 1 | mineruDocument parser that converts PDFs, images and Office files to Markdown and JSON | 81 out of 100 |
| 2 | Crawl4AIPython web crawler that converts pages to LLM-ready Markdown and JSON | 79 out of 100 |
| 3 | doclingPython library and CLI that converts PDFs, Office files and more into structured text | 79 out of 100 |
| #4 | label-studioMulti-user data labeling tool for audio, text, images, video and time series | 74 out of 100 |
| #5 | markitdownPython tool that converts PDF, Office, HTML and audio files to Markdown for LLMs | 71 out of 100 |