Skip to content

Open-source RAG platform · Cloud · On-premises · Air-gapped

Private AI assistant for your company's documents

DocsGPT is an open-source platform for private AI agents, assistants and enterprise search over an organisation's own documents, built by Arc53 and deployable in the cloud, on your own servers or fully air-gapped. Every answer is cited to its source; the licence is MIT.

curl -fsSL https://docs.ac/install | bash

One command, Docker requiredQuickstart

zsh — docsgpt

Used by people from

  • Microsoft
  • Intel
  • Amazon
  • Airbnb
  • Siemens
  • Deloitte
  • PwC
  • Accenture
  • AstraZeneca
  • HM Government
  • Payoneer
  • Stanford
  • MIT
  • DWP

    UK Department for Work and Pensions: an internal knowledge assistant on government infrastructure, with local inference.

    Public sector · local inference

  • AMD

    AMD runs DocsGPT as an assistant over internal technical documentation.

    Internal technical documentation

  • LexEU

    LexEU: multilingual EU-law research over 2.5M+ pages, answers in 6 EU languages with full source citations.

    2.5M+ pages · 6 EU languages · cited

Run it where your data is allowed to be: Cloud · Managed · On-premises · Air-gapped

Same platform, same MIT licence, from one seat. Hybrid (on-premises platform, your cloud LLM) is an On-premises option; air-gapped means zero external network. Or self-host the community edition.

Who operates each posture, where documents are stored and where the language model runs. Cloud and Managed are operated by Arc53; On-premises and Air-gapped run in your own environment with Arc53 support.

Deployment postures: operator, data location, LLM location
 CloudManagedOn-premisesAir-gapped
OperatorArc53, multi-tenantArc53, dedicated instanceCustomer, with Arc53 supportCustomer, with Arc53 support
Data locationArc53 cloud, EU regionDedicated VPS in the Americas, Europe, Middle East, Africa or AsiaYour data centre or private cloudYour network, no external egress
LLM locationCloud provider via Arc53Cloud provider, local engine or your own keyLocal engine, or your cloud LLM (hybrid)Local engine only
External egressYesYesOptionalNone
Details/deploy/cloud →/deploy/managed →/deploy/on-premises →/deploy/air-gapped →

Hybrid (on-premises platform + cloud LLM) is listed under On-premises.

One self-hosted RAG platform: knowledge, agents, workflows, governance

Four parts of one open-source codebase.

Knowledge → Agents → Workflows → Governance
  1. 01Knowledgeingest · OCR · index
  2. 02Agentsretrieve · cite · act
  3. 03Workflowsroute · check · approve
  4. 04Governanceguardrails · audit · SSO
Text version

Documents are connected, parsed and indexed (Knowledge). Agents answer over them with citations and call tools (Agents). Multi-step jobs run node by node with human approval where a step is uncertain (Workflows). Guardrails, the audit log and SSO/SCIM/RBAC apply across all of it (Governance).

Knowledge

Connect and index your documents

30+ file formats, connectors for Google Drive, SharePoint, Confluence and S3, OCR for scanned PDFs, chunk management per source. Docs: sources

pdf · docx · xlsx · pptx · md · html · drive · sharepoint · confluence · s3

Agents

Agents that answer and act

Classic, Agentic, Research and Workflow agent types; scheduled runs; tools over MCP and JSON-schema APIs. Docs: agents

classic · agentic · research · workflow · cron · mcp

Governance

Guardrails, approvals, audit, SSO

7 guardrail checks that flag, redact or block; human approval for answers and tool calls; append-only audit log; OIDC SSO, SCIM 2.0 and RBAC via Admin. Docs: guardrails

guardrails · approvals · audit-log · oidc · scim · rbac

Results from deployments

24 h → 16 min

A large Central European commercial bank processes 10,000+ inbound documents a day; time-to-action fell from 24 hours to 16 minutes, with over 99.5 % classification accuracy, running fully air-gapped on the bank's existing NVIDIA T4 GPUs.

Source: EU bank, name under NDA

Government infrastructure, local inference

UK Department for Work and Pensions: an internal knowledge assistant on government infrastructure, with local inference.

Source: DWP

2.5M+ pages · 6 languages

Multilingual EU-law research over 2.5M+ pages, answers in 6 EU languages with full source citations, 100–200 ms retrieval.

Source: LexEU

Built for regulated environments

Identity, audit, guardrails and open code. Threat model and disclosure process on the security page.

Identity

OIDC single sign-on, SCIM 2.0 provisioning and role-based access control.

oidc · scim 2.0 · rbac

Audit

An append-only audit log records what was asked, answered and executed.

audit-log · append-only

Guardrails

Checks on inputs and outputs that flag, redact or block, with human approval gates.

flag · redact · block

Open code

MIT-licensed source, a public threat model, an SBOM and CodeQL scanning on every change.

MIT · THREAT_MODEL.md · sbom · codeql

Security posture, threat model and disclosure →

Open source, MIT licence

The whole platform is one public repository. Read it, run it, contribute to it.

GitHub stars

18,274

arc53/DocsGPT

Contributors

226+

on GitHub

Licence

MIT

run and modify

Frequently asked questions

Are there any AI chatbots that are private?

Yes. DocsGPT runs on your servers or air-gapped with local models, so no document or question leaves your network; the code is MIT-licensed and inspectable.

Can AI be on prem?

Yes. DocsGPT deploys with Docker Compose or Kubernetes manifests in your data centre and runs local engines such as Ollama, vLLM or llama.cpp; a bank runs it fully air-gapped on existing NVIDIA T4 GPUs.

Is there a self-hosted AI that cites its sources?

Every DocsGPT answer carries title, source and page number, and retrieval can be hybrid (BM25 + vector) or GraphRAG.

What does it cost?

Cloud is $20/seat/month; Managed is from $2,000/month, flat; On-premises and Air-gapped are from $10,000, one-time or recurring — per deployment, not per seat. Individuals can start free in the app. See pricing.

See DocsGPT on your documents

A 30-minute demo on your use case, or start free in the cloud app. Self-hosting takes one command.

Last updated: