evalpair for Grok
io.github.srikarmk/evalpair
Run LLM eval suites on local or hosted models; compare runs with intervals and a noise floor.
client:Grok
transport:stdio
runtime:pypi
Install evalpair in Grok
.mcp.json (in your project root)
{
"mcpServers": {
"evalpair": {
"type": "stdio",
"command": "uvx",
"args": [
"evalpair"
],
"env": {
"EVALPAIR_HOME": "<EVALPAIR_HOME>",
"EVALPAIR_API_KEY": "<EVALPAIR_API_KEY>"
}
}
}
}
Grok Build reads Claude Code's .mcp.json and Claude Desktop's config unchanged, so a server you have already set up there needs nothing new. The Grok web app and X integration do not take custom servers. Grok docs