Skip to content

Comparison ​

lockdocsContext7docs-mcp-serverRef
Where docs come fromYour installed packagesHosted index of public repos and sitesSites and repos you scrape into a local indexHosted index
Which versionExact, from your lockfileChosen by library ID or prompt; a few indexed versions per libraryThe version you scrapedMostly latest
Per-library setupNoneNoneScrape each library (and version)None
Private / internal packagesYes, if installedPaid plansYes, if you scrape themPrivate GitHub repos and PDFs you connect (plan limits)
API reference from types and doc comments.d.ts, docstrings, rustdoc, Go docDoc snippetsDoc pagesDoc pages
Works offlineYesNoAfter scrapingNo
LimitsNoneAnonymous ratelimit-limit: 200 (observed); free key 1,000 calls/monthNone200 free credits once, then from $19/month
Account / API keyNoOptionalOptional (embeddings)Yes
LicenseMITMIT client, hosted serviceMITProprietary

Context7 knows many sites and guides that never ship inside a package, so for a question whose answer lives only on a docs website, it can do better. lockdocs is for the other case: when the version matters, when you are offline or rate-limited, and when the package is yours. They combine well.

When to use which ​

  • lockdocs: the version matters (a major upgrade, a pinned older release), you are offline or rate-limited, the package is private or internal, or you want the exact signature from the installed types.
  • Context7 / Ref: the answer lives in a guide on a docs website that is not shipped inside the package, or the library is not installed.
  • docs-mcp-server: you want a local semantic index of specific websites and are happy to scrape each one.

Running lockdocs next to a hosted doc server is a good setup: lockdocs answers "what does my version do", the hosted one fills in website-only guides.

MIT licensed · local, offline, no API key