Unstract
No-code LLM platform that turns invoices, forms and other documents into structured JSON.
Quick answer
Unstract is a no-code LLM platform that turns invoices, forms and other documents into structured JSON, listed under Document processing. A typical small-business use: extract fields from invoices, purchase orders, bank statements or insurance forms into JSON. Official hardware: minimum 8 GB RAM. That fits a Hetzner Cloud CPX32 server at about €35.99/month (server only, excl. VAT). License: AGPL-3.0, which allows internal business use.
Last updated · facts verified against official docs on · sources
Activity data comes from GitHub and was last refreshed Oct 26, 2026.
What is Unstract?
Unstract lets you define what to extract from a document type with natural-language prompts in Prompt Studio, then run that schema as a REST API, an ETL pipeline from cloud storage into a database or warehouse, an MCP server or an n8n node. It works with OpenAI, Anthropic, Bedrock, Ollama and other LLM providers plus your choice of vector database and text extractor.
Where it fits in a 10–200 person business
- Extract fields from invoices, purchase orders, bank statements or insurance forms into JSON
- Batch-process a folder or S3 bucket of documents into PostgreSQL, Snowflake or BigQuery
- Add a document-to-JSON API step to n8n or other automation workflows
These use cases are our editorial suggestions, not claims made by the project.
Can a business use Unstract? (license)
| License | AGPL-3.0 |
|---|---|
| Type | Network copyleft (AGPL): Like GPL, and also applies when users access a modified version over a network. |
| Internal business use | Allowed under the license text |
| What to watch | Free for internal and commercial use. If you modify Unstract and let users interact with it over a network, you must offer them the modified source code under AGPL-3.0. |
This is a plain-English summary of the license text, not legal advice. Third-party models you run with it have their own licenses.
Data & compliance: does Unstract phone home?
| Telemetry (usage data sent to the vendor) | On by default The frontend integrates PostHog to track minimal usage analytics by default. How to turn it off: Set REACT_APP_ENABLE_POSTHOG=false in the frontend .env file. |
|---|---|
| Other calls home by default (updates, marketplaces, default cloud services) | Not documented |
| Single sign-on (OIDC / SAML / LDAP) | Paid / enterprise only By protocol: OIDC paid · SAML paid. SAML/OIDC SSO is listed under Unstract Cloud & Enterprise. |
| Audit log | Not documented |
| Role-based access control (RBAC) | Not documented Granular enterprise RBAC is listed as a Cloud & Enterprise feature; the docs do not say which roles the open-source edition has. |
| Data residency / GDPR notes from the project | Not documented |
Checked against the project's own docs and repository on . "Not documented" means we could not find the answer in official sources; it does not mean the feature is missing or that no data is sent. Third-party services you connect (model APIs, vector databases, search engines) have their own data policies. Compare all apps.
What hardware does Unstract need?
| Minimum | 8 GB RAM |
|---|---|
| Recommended | Not published |
| GPU | Not required by the platform itself; LLMs are reached through the provider you configure (cloud API or a local server such as Ollama). |
8 GB RAM minimum on Linux or macOS (Intel or M-series), with Docker and Docker Compose. No CPU or disk figures are published.
How much does it cost to host Unstract per month?
| Minimum spec | 8 GB RAM → CPX32 (4 vCPU, 8 GB, 160 GB disk) €35.99/mo |
|---|
How we got this: we took the cheapest Hetzner Cloud Regular Performance (shared AMD vCPU) plan with at least the RAM and vCPU the project lists, using prices read from www.hetzner.com/cloud/regular-performance on 2026-10-26. "From" monthly price cap shown by Hetzner, excluding VAT. Prices vary by location. Covers the server only; it excludes LLM API usage, backups, extra storage, GPUs and your time. Disk is not matched; add a volume if you need more than the plan includes.
How to deploy Unstract
Method: Docker Compose (run-platform.sh)
git clone https://github.com/Zipstack/unstract.git
cd unstract
./run-platform.sh
# pin a release: ./run-platform.sh -v <version>; upgrade: ./run-platform.sh -uFrom the official instructions: github.com/Zipstack/unstract#readme. Always check the source for the current version before you run it.
Known gotchas when self-hosting Unstract
- The default login is unstract / unstract. Set DEFAULT_AUTH_USERNAME and DEFAULT_AUTH_PASSWORD in backend/.env and restart before exposing it. source
- PostHog usage analytics are on by default in the frontend. Set REACT_APP_ENABLE_POSTHOG=false in the frontend .env to disable them. source
- SSO (SAML/OIDC), granular enterprise RBAC, human-in-the-loop review, LLMChallenge dual-LLM verification and SinglePass extraction are Cloud/Enterprise features, not part of the open-source platform. source
- The quickstart lists Linux or macOS only, and needs at least 8 GB RAM for the full Docker stack. source
Alternatives in Document processing
- PaddleOCR OCR toolkit and document AI engine that turns scans, images and PDFs into LLM-ready text, Markdown or JSON. Compare PaddleOCR vs Unstract
- MinerU Document parser that converts PDFs, scans and Office files into clean Markdown or JSON for LLMs and RAG. Compare MinerU vs Unstract
- Paperless-AI AI auto-tagging and document chat for Paperless-ngx (currently unmaintained). Compare Paperless-AI vs Unstract
Unstract FAQ
How much RAM does Unstract need?
The official minimum is 8 GB RAM. 8 GB RAM minimum on Linux or macOS (Intel or M-series), with Docker and Docker Compose. No CPU or disk figures are published.
Does Unstract need a GPU?
Not required by the platform itself; LLMs are reached through the provider you configure (cloud API or a local server such as Ollama).
How much does it cost to self-host Unstract?
The software is free to download. Based on the official minimum 8 GB RAM: That fits a Hetzner Cloud CPX32 server at about €35.99/month (server only, excl. VAT). That covers the server only, not LLM API usage, backups, extra storage or your time.
Is Unstract free for commercial use?
Unstract is licensed under AGPL-3.0 (Network copyleft (AGPL)). Internal business use is allowed under the license text. Free for internal and commercial use. If you modify Unstract and let users interact with it over a network, you must offer them the modified source code under AGPL-3.0. This is a summary, not legal advice.
What are common problems when self-hosting Unstract?
The default login is unstract / unstract. Set DEFAULT_AUTH_USERNAME and DEFAULT_AUTH_PASSWORD in backend/.env and restart before exposing it. PostHog usage analytics are on by default in the frontend. Set REACT_APP_ENABLE_POSTHOG=false in the frontend .env to disable them. SSO (SAML/OIDC), granular enterprise RBAC, human-in-the-loop review, LLMChallenge dual-LLM verification and SinglePass extraction are Cloud/Enterprise features, not part of the open-source platform. The quickstart lists Linux or macOS only, and needs at least 8 GB RAM for the full Docker stack.
Is Unstract still maintained?
Yes, it is actively developed. Last commit: Oct 26, 2026. Latest release: v0.197.0 (Oct 25, 2026). Checked Oct 26, 2026.
How do I install Unstract?
The official method is Docker Compose (run-platform.sh). The exact commands are in the "How to deploy" section above and in the official instructions: https://github.com/Zipstack/unstract#readme
What are self-hosted alternatives to Unstract?
Other apps in Document processing: PaddleOCR, MinerU, Paperless-AI.
Sources
- GitHub repository: github.com/Zipstack/unstract
- Official documentation: docs.unstract.com/unstract/unstract_platform/quick_start
- License text: github.com/Zipstack/unstract/blob/main/LICENSE
- Hardware requirements, Deploy instructions, Gotcha 2, Gotcha 3, Gotcha 4: github.com/Zipstack/unstract#readme
- Gotcha 1: github.com/Zipstack/unstract/blob/main/backend/README.md
- Hetzner Cloud pricing (cost estimate): www.hetzner.com/cloud/regular-performance
- GitHub stars, commits and releases: GitHub REST API, fetched
Facts on this page were checked against the project's own repository and docs on 2026-10-26. Spotted something out of date? Tell us.