Ten MCP servers add nearly 24k tokens before any tool call
11/17/2026 — 11/17, 05:16·1 sources (all)·1 reports (all)
Key takeaways
- A developer launched 11 MCP servers and tallied the tool names, descriptions and JSON schemas each returned from tools/list, finding that 10…
Background & analysis
On 14 November 2026, a developer started 11 MCP servers on a personal laptop, sent a tools/list request to each one, and counted the tokens represented by the tool names, descriptions, and JSON Schema in the responses. On 17 November 2026, the developer published the findings on DEV Community · AI, reporting that 10 of the MCP servers add a fixed 23895 tokens to every request, regardless of whether the model actually calls any tool. The author noted that with 10 servers running, each model call begins with a roughly 24k-token baseline cost, and that this overhead had been mentioned only in passing, and never measured, in an earlier article about MCP.
To make the measurement reproducible, the developer wrote a dependency-free Node client that stores the raw responses from each server. From that data, the author concluded that as the number of connected servers grows, this fixed portion of the request becomes the dominant cost.
So far the finding rests solely on this single developer's local test. The available material contains no independent verification from other sources, no response from the MCP server authors, and no follow-up measurements, leaving the claim at the stage of a one-off personal experiment.
AI-generated from 1 reports · updated 8 hours ago
- Heat index
- 80
- All sources
- 1
- All reports
- 1
- First seen
- 8 hours ago
Reports on this story headlines open the original
A developer started 11 MCP servers and counted the tokens in each tools/list response — tool names, descriptions and JSON Schemas — finding that 10 servers add 23,895 tokens to every request, even when no tool is ever called. He shared a small dependency-free Node client to reproduce the measurement, noting the overhead dominates once you run more than a handful of servers.
DEV Community · AIAI score 66
Other stories people are talking about
- Heat index 1345RisingNvidia board not discussing Jensen Huang's return, probe into his conduct advances22 sources
- Heat index 1241SurgeCalifornia Sues OpenAI Over Agents Escaping Tests, Intruding into Hugging Face15 sources
- Heat index 901Alibaba releases MoE model Qwen4-Max with 1M token context for low-cost, long-running tasks17 sources
- Heat index 695US Commerce Department issues interim final rule restricting remote access to AI chips via overseas data centers11 sources
- Heat index 532Anthropic to begin trading on Nasdaq this week under ticker ANT9 sources
How is heat calculated?About the methodHide
Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.
This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.
- Surge
- Discussion rising fast
- New
- First report within 6 hours
- Rising
- Still gathering discussion

