evalpair
Run LLM eval suites on local or hosted models; compare runs with intervals and a noise floor.
Install
One-Click Install
Add this to your `claude_desktop_config.json` file:
{
"mcpServers": {
"evalpair": {
"command": "uvx",
"args": [
"evalpair"
]
}
}
}Runnable packages
pypievalpair
0.1.1stdioWhat it can do
Tool inventory
No publishable tool enumeration has been recorded.
Community
Evidence
Recent observations
This server has Registry metadata but no public verification run yet.