mcpcost

io.github.moezdil/mcp-gpu-server

GPU Server

Is the GPU Server MCP server down? No. It answered initialize and tools/list in the probe of 2026-10-27.

MCP server for GPU monitoring: nvidia-smi, VRAM, utilization, temperature

alive probed 2026-11-15 · unchanged since 2026-10-27

5tools
397tokens in Claude Opus 5.5
0.2 %of Claude's 200k window
0.2×the median server
1.6 sto initialize
5 msto list tools
Run it
uvx gpu-server@0.2.2 (stdio, probed in a container with the network off)
Host
pypi.org
Reports itself as
mcp-gpu-server
Capabilities
prompts, resources, tools
Instructions
none
Of a 1M window
<0.1 % of a 1M context window
Count
397 tokens in Claude Opus 5.5, exact, measured 2026-10-27. With the Claude Code prefix mcp__…__ on each tool name: 467. The o200k_base estimate is 195. Counted again when the tool definitions change. Method.
Repository
https://github.com/moezdil/mcp-gpu-server
Website
https://github.com/moezdil/mcp-gpu-server
Registry
active, updated 2026-10-21 · registry entry

Most expensive tools

The Claude column scales the o200k estimate of each tool by this server's ratio of exact to estimate, 2.04×.

Tool≈ Claude tokenso200k estimateShare of this server
gpu_vram924523 %
gpu_stats834121 %
gpu_info773819 %
gpu_utilization753719 %
gpu_temperature693417 %

Compare this server with others, or with the servers that share its tool set.

Same tool set as 1 other server

These registry entries have the same tool names and the same schema sizes.

More servers

Other registry entries on pypi.org, the ones next to this name; all of them.

Servers that mention temperature or utilization.

Badge for the README

mcpcost: alive

The result of the latest probe, the tool count and the token cost. The feed has an entry for every probe that changed them.

Atom feed

A mail when this server changes

Stops answering, comes back, tool set or cost changes. One confirmation mail first, a stop link in every mail.

For the publisher