Tome, an MIT-licensed documentation platform at github.com/vxcozy/tome, ships a Model Context Protocol server out of the box — run `tome mcp` and your docs are immediately consumable by AI tools and agents. That alone separates it from every major free alternative. The platform also includes an embedded AI chat assistant with bring-your-own-key support for OpenAI and Anthropic models, so teams can add Q&A to their docs without routing through a third-party subscription. Both features are built-in, not plugins bolted on after the fact.
On cost, Tome targets developers frustrated with Mintlify's $300-plus monthly pricing. Self-hosting is free; the managed cloud tier runs $19 per month. Setup is a single `npx create-tome` command. At 237 GitHub stars and over 111 CI/CD workflow runs at time of writing, it appears to be picking up early traction from teams looking to cut documentation infrastructure costs without giving up polish.
The feature set is broad for a new entrant: Markdown and MDX authoring, Shiki syntax highlighting, OpenAPI-based interactive API references, Pagefind local search alongside Algolia DocSearch, i18n, versioning, privacy-first analytics, and automated OG image generation. One-command migration tooling for both GitBook (`tome migrate gitbook`) and Mintlify (`tome migrate mintlify`) makes the acquisition strategy explicit. The codebase is a monorepo with discrete packages for core processing, CLI, theming, and MDX components; a plugin system with custom build hooks and Vite plugin support suggests an intent to grow a third-party ecosystem.
Against Docusaurus — the dominant free alternative maintained by Meta — Tome competes primarily on developer experience, advertising roughly a two-minute setup versus Docusaurus's estimated thirty minutes for equivalent functionality. Docusaurus requires plugin assembly for OpenAPI references and search, areas where Tome ships built-in solutions. Neither Docusaurus nor free alternatives like Nextra or VitePress offer a managed hosting tier or <a href="/news/2026-03-14-optimizing-web-content-for-ai-agents-via-http-content-negotiation">an MCP server</a> as documented core features. That gap is where Tome is planting its flag.