EvalForge Lite runs on your own machine — the client starts it, so there is no endpoint to ping. 122 installs a week from pypi. Last commit 15 Oct 2026.
Compare text LLMs across OpenRouter, Bedrock, Vertex AI and Foundry with automated grading.
Today is the operative word: we check EvalForge Lite every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add evalforge-lite -- uvx evalforge-lite
{
"mcpServers": {
"evalforge-lite": {
"command": "uvx",
"args": [
"evalforge-lite"
]
}
}
}
[mcp_servers.evalforge-lite]
command = "uvx"
args = ["evalforge-lite"]
{
"mcpServers": {
"evalforge-lite": {
"command": "uvx",
"args": [
"evalforge-lite"
]
}
}
}
{
"servers": {
"evalforge-lite": {
"type": "stdio",
"command": "uvx",
"args": [
"evalforge-lite"
]
}
}
}
Optional settings from the registry entry, left out of the block above:
OPENROUTER_API_KEY (Optional. OpenRouter API key held by the server so tool calls need no credentials for OpenRouter. If unset, pass creds in each tool call).
Route coding work across Claude Code, Codex, Cursor, and OpenCode with shared context.
Secure durable memory and bounded context for AI agents, with hosted OAuth and local Lite support.
Chat with 300+ LLMs via OpenRouter. Analyze and generate images, audio, and video from MCP.
Let AI assistants shop for home and auto insurance: compare quote estimates across every carrier.
Compare AI model prices and estimate workload costs with four read-only tools. No account needed.
Persistent memory and shared context across compatible AI assistants, with account-scoped OAuth.
Ask, compare and judge 33 Bedrock text models, generate and edit images, plus video and embeddings.
Competitors' Google ads for AI agents: every ad a brand runs, with dates and ad text.
Answers built from our own checks of this server.