DeepDoc
deepdoc.tech
- Free plan
- Windows
- macOS
- Linux
- Web
- Self-hosted
What DeepDoc is
DeepDoc turns a code repository into a navigable documentation site and can regenerate pages affected by code changes. It scans source files, plans documentation, generates pages, stages OpenAPI references when present, and builds a site. Documentation can be organized into topics such as architecture, endpoints, integrations, and schemas. Its AI answers can cite repository file paths and line ranges. DeepDoc supports Python, TypeScript, JavaScript, Go, PHP, Vue, GraphQL, and OpenAPI or Swagger specifications. The maker describes a terminal-based tool that needs no account or cloud service; it is an MIT-licensed Python package requiring Python 3.10 or later. Users must supply an LLM provider. The generate and update commands use that provider, with costs depending on the model and documentation pages; init, serve, deploy, status, config, performance, and clean do not call an LLM. Output options listed include a MkDocs Material site, Markdown, Mermaid diagrams, a static Next.js/Fumadocs site, and an interactive API reference. The deploy command produces a static site for any host.
Who it suits
DeepDoc suits developers who want repository-based documentation and can configure a Python environment and provide an LLM provider. Its local operation may also fit teams that do not want an account or cloud requirement for the tool.
Good to know, and what to watch
- Generates documentation organized by code topics
- AI answers can cite source paths and line ranges
- Incremental update regenerates affected pages
- Runs in the terminal without an account or cloud requirement
- Static site can be deployed to any host
- Requires Python 3.10 or later
- Users must bring an LLM provider
- Generate and update incur provider-dependent costs
Plans
Every plan and priceDeepDoc
Free plan
The verdict
DeepDoc offers a local, MIT-licensed way to turn repositories into structured documentation sites. Its generation workflow depends on a separately chosen LLM, so account for that provider’s costs when planning updates.
What the maker publishes
- Company details
- The pages opened identify no headquarters or founding year; the homepage says DeepDoc is not backed by Y Combinator. source
- Deployment
- DeepDoc produces a static site that can be deployed to any host. source
- Documentation structure
- It groups documentation into topics such as architecture, endpoints, integrations, and schemas. source
- Generation pipeline
- DeepDoc scans source files, plans documentation, generates pages, stages OpenAPI references when present, and builds a site. source
- Hosted output
- The docs say deepdoc deploy produces a static site for any host. source
- Incremental updates
- The update command regenerates pages affected by code changes. source
- License
- The homepage identifies DeepDoc as MIT licensed. source
- License and runtime
- The homepage identifies DeepDoc as MIT licensed and requiring Python 3.10 or later. source
- LLM requirement
- DeepDoc requires users to bring their own LLM provider. source
- LLM setup
- DeepDoc requires users to bring their own LLM provider. source
- Local operation
- The maker says DeepDoc runs in the terminal with no account or cloud requirement. source
- Plans and pricing
- The maker's cost page describes usage-based LLM costs and provides no dollar estimates; it does not list subscription plans. source
- Pricing model
- The maker says generation and update call an LLM, while commands such as serve and deploy do not; it gives no dollar price because usage depends on the model and planned page count. source
- Product format
- The maker provides a Python package installed with pip install deepdoc. source
- Purpose
- DeepDoc turns a code repository into a navigable documentation site and keeps the docs current as code changes. source
- Site output
- DeepDoc generates a MkDocs Material site with navigation, search, and an optional AI chatbot widget. source
- Source-backed answers
- Its AI answers cite source file paths and line ranges from the codebase. source
- Source citations
- Its AI answers cite source file paths and line ranges from the repository. source
- Supported inputs
- The homepage lists Python, TypeScript, JavaScript, Go, PHP, Vue, and OpenAPI as supported. source
- Supported languages
- The homepage lists Python, TypeScript, JavaScript, Go, PHP, Vue, and OpenAPI as supported. source
- Usage costs
- The maker says init, serve, deploy, status, config, performance, and clean do not call an LLM; generate and update do, so costs depend on the chosen model and documentation pages. source
Checked against the maker's pages on 7 Oct 2026.