Skip to content
mcp/skillhub

Compare · Documentation

Context7 vs Microsoft Learn MCP Server

Context7 exposes 2 tools and adds about 1,188 tokens to every model turn. Microsoft Learn exposes 3 tools and adds about 1,016. Measured by connecting to each server and counting its tool list, not from the README.

Context7upstashMicrosoft Learnmicrosoftdocs
Context tax~1,188 tok/turn~1,016 tok/turn(leaner)
Tools23
Cost per 1,000 turns (Claude Sonnet 5.5, uncached)$2.38$2.03
Security score100/100100/100
Cold start——
RunsLocal or hostedLocal or hosted
Needs keysCONTEXT7_API_KEYNone declared
FreshnessActiveMaintained
GitHub stars63k2.0k
LicenseMITCC-BY-4.0
Official MCP registryYesYes

Install in Claude Code

Context7

claude mcp add context7 -e CONTEXT7_API_KEY='<context7-api-key>' -- npx -y @upstash/context7-mcp
Cursor, VS Code, Claude Desktop, Windsurf

Microsoft Learn

claude mcp add microsoftdocs -- npx -y @microsoft/learn-cli
Cursor, VS Code, Claude Desktop, Windsurf

Tools

By name, from each server's live tool list. The same name can still behave differently.

Only in Context7 · 2

  • resolve-library-id
  • query-docs

In both · 0

  • none

Only in Microsoft Learn · 3

  • microsoft_docs_search
  • microsoft_code_sample_search
  • microsoft_docs_fetch

Questions

Which uses fewer tokens, Context7 or Microsoft Learn?
Microsoft Learn: its tool definitions take about 1,016 tokens per turn against 1,188 for Context7 (1.2× more). Tool definitions are resent on every model turn.
Do Context7 and Microsoft Learn have the same tools?
No tool names overlap: Context7 has 2 tools and Microsoft Learn has 3.
Are Context7 and Microsoft Learn hosted or local?
Context7: local or hosted. Microsoft Learn: local or hosted.

Token counts use the o200k tokenizer on each server's tool definitions as an agent receives them; clients that load tools lazily pay less. Prices are Claude Sonnet 5.5 input at $2/MTok. Want to check one yourself? Run npx mcpskillhub test <package> — how it works.