Llmtest MCP runs on your own machine — the client starts it, so there is no endpoint to ping. 39 installs a week from npm. Last commit 22 May 2026.
Benchmark AI models on real prompts. Find cheaper, faster alternatives across 340+ models.
Today is the operative word: we check Llmtest MCP 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 llmtest-mcp -- npx -y llmtest-mcp
{
"mcpServers": {
"llmtest-mcp": {
"args": [
"-y",
"llmtest-mcp"
],
"command": "npx"
}
}
}
[mcp_servers.llmtest-mcp]
command = "npx"
args = ["-y", "llmtest-mcp"]
{
"mcpServers": {
"llmtest-mcp": {
"args": [
"-y",
"llmtest-mcp"
],
"command": "npx"
}
}
}
{
"mcpServers": {
"llmtest-mcp": {
"args": [
"-y",
"llmtest-mcp"
],
"command": "npx"
}
}
}
This one needs environment variables set before it will start:
LLMTEST_API_KEY (Your LLMTest API key from https://llmtest.io/dashboard).
The author declared them in the registry entry; get the values from the project itself.
Price benchmarks, alternatives & daily price history across 17,000+ AI agents and MCP servers.
A second opinion for AI agents: one prompt across several live Gonka models + roles, one call.
Live AI compute pricing oracle — real-time LLM model prices across providers
Real-time LLM/VLM benchmarks, pricing, and recommendations. 300+ models, 5 sources.
Query AI model benchmarks, pricing, and comparisons from rival.tips
Private-market data across Southeast Asia via the Alternatives Partner API v3
Free AI models and per-token prices across API gateways. Search, compare, estimate cost. Read-only.
AI Model Experiments MCP ('Model Lab') — run the same prompts across many
Answers built from our own checks of this server.