Skip to content
mcp/skillhub

Compare · Web Search

Fetch vs Firecrawl MCP Server

Fetch exposes 1 tools and adds about 235 tokens to every model turn. Firecrawl exposes 3 tools and adds about 4,383. Firecrawl costs 18.7× as much context. Measured by connecting to each server and counting its tool list, not from the README.

FetchmodelcontextprotocolFirecrawlfirecrawl
Context tax~235 tok/turn(leaner)~4,383 tok/turn
Tools13
Cost per 1,000 turns (Claude Sonnet 5.5, uncached)$0.47$8.77
Security score95/10097/100
Cold start3,873 ms—
RunsLocal (stdio)Local or hosted
Needs keysNone declaredFIRECRAWL_API_KEY
FreshnessActiveActive
GitHub stars91k7.6k
License—MIT
Official MCP registryNoYes

Install in Claude Code

Fetch

claude mcp add fetch -- uvx mcp-server-fetch
Cursor, VS Code, Claude Desktop, Windsurf

Firecrawl

claude mcp add firecrawl -e FIRECRAWL_API_KEY='<firecrawl-api-key>' -- npx -y firecrawl-mcp
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 Fetch · 1

  • fetch

In both · 0

  • none

Only in Firecrawl · 3

  • firecrawl_scrape
  • firecrawl_search
  • firecrawl_parse

Questions

Which uses fewer tokens, Fetch or Firecrawl?
Fetch: its tool definitions take about 235 tokens per turn against 4,383 for Firecrawl (18.7× more). Tool definitions are resent on every model turn.
Do Fetch and Firecrawl have the same tools?
No tool names overlap: Fetch has 1 tools and Firecrawl has 3.
Are Fetch and Firecrawl hosted or local?
Fetch: local (stdio). Firecrawl: 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.