modelcostsaver for Grok
io.github.sachinuppal/modelcostsaver
Offline MCP: predict LLM call cost and pick the cheapest capable model. No keys, no network.
client:Grok
transport:stdio
runtime:npm
Install modelcostsaver in Grok
.mcp.json (in your project root)
{
"mcpServers": {
"modelcostsaver": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@workswarm/modelcostsaver"
]
}
}
}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