An auto-generated wiki for all your code
DeepWiki Open analyzes your GitHub, GitLab, or BitBucket repositories and produces an interactive wiki, diagrams, and an assistant that answers your questions about the code. In local mode, your code never leaves our servers in France.
What is DeepWiki Open?
DeepWiki Open automatically generates interactive wikis and documentation for any GitHub, GitLab, or BitBucket repository. The tool clones the repository, analyzes the code structure, produces structured documentation, Mermaid diagrams, and a navigable wiki. An 'Ask' feature allows you to query the code in natural language using semantic RAG search.
The architecture is based on a Python backend (cloning, traversal, and embeddings pipeline) and a Next.js / TypeScript interface. Code embeddings are indexed in a local FAISS vector store; persistence is file-based, with no relational database server to manage.
Key privacy feature: DeepWiki Open is multi-provider and can run 100% locally via Ollama (embeddings and generation), with no external API keys required. It also accepts, at your choice, OpenAI, Google Gemini, Anthropic, or OpenRouter APIs. Deployment is done via Docker, with a dedicated Dockerfile for the local Ollama mode.
Host DeepWiki Open at DINAO
Resource tiers compatible with DeepWiki Open prerequisites (minimum 2 vCPU / 2 Go / 5 Go). Hosted in France, fully managed.
- 2 dedicated vCPU
- 4 Go RAM
- 40 GB NVMe
- Daily backups
- Managed & monitored by DINAO
- 4 dedicated vCPU
- 8 Go RAM
- 80 GB NVMe
- Daily backups
- Managed & monitored by DINAO
- 8 dedicated vCPU
- 16 Go RAM
- 160 GB NVMe
- Daily backups
- Managed & monitored by DINAO
1 hidden tier(s) (insufficient resources for this app) : Découverte
This application uses AI
The container hosts the application, not the AI engine (which requires dedicated GPUs) : inference is handled externally, with your own provider key. Prioritize a sovereign engine — Mistral AI, NumSpot, Scaleway, or OVHcloud AI Endpoints (France, GDPR) ; an international provider (OpenAI, Anthropic…) only if a specific capability requires it. Inference subscriptions are not included in hosting.
Technical details
You might be wondering…
Does my source code go to a third party?
Not if you choose local mode: connected to open-weight models via Ollama, DeepWiki Open generates embeddings and documentation entirely on your instance, with no external API keys. In external API mode (OpenAI, Gemini…), the code is transmitted to the provider — to be avoided for sensitive code.
Which repositories can I document?
Any GitHub, GitLab, or BitBucket repository, public or private. DeepWiki clones the repository, analyzes its structure, and produces a navigable wiki with diagrams.
What is the 'Ask' feature?
A RAG-type assistant: you ask a question in natural language about the code ('how does payment work?') and it answers based on actual files, using semantic search via a local FAISS vector index.
Where are the data hosted?
On DINAO infrastructure in France, in one of the available data centers. In local mode, code, embeddings, and generated wikis remain on the server.
Do I need a GPU?
No to start: small models run on CPU. For large repositories or more powerful local models, a GPU (optional or dedicated on higher tiers) significantly speeds up generation.
