I measured the token cost of 9 MCP servers: 41k tokens before the first prompt
This story is from 2026-09-22. It is preserved in the archive; the latest stories are on the live feed.
Every MCP server you connect costs tokens before you type anything. The server's tool definitions get injected into the model's context, and the model reads them on every single call. I wired up 9 servers over a few weeks, never counted the total, and then one day I ran /context and found out what…
Read the full story at DEV Community — AI ↗
Timeline · 1 report
- 2026-09-22 06:07 · DEV Community — AI
I measured the token cost of 9 MCP servers: 41k tokens before the first prompt