LLM Advisor runs on your own machine — the client starts it, so there is no endpoint to ping. 156 installs a week from npm. Last commit 18 Jun 2026.
Real-time LLM/VLM benchmarks, pricing, and recommendations. 300+ models, 5 sources.
Nothing serious here today
Today is the operative word: we check LLM Advisor every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
Connect this server
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 llm-advisor -- npx -y llm-advisor-mcp{
"mcpServers": {
"llm-advisor": {
"args": [
"-y",
"llm-advisor-mcp"
],
"command": "npx"
}
}
}[mcp_servers.llm-advisor]
command = "npx"
args = ["-y", "llm-advisor-mcp"]{
"mcpServers": {
"llm-advisor": {
"args": [
"-y",
"llm-advisor-mcp"
],
"command": "npx"
}
}
}{
"mcpServers": {
"llm-advisor": {
"args": [
"-y",
"llm-advisor-mcp"
],
"command": "npx"
}
}
}Alternatives to LLM Advisor
same job, measured the same wayLLM provider intelligence: recommendations, pricing, and status
Query AI model benchmarks, pricing, and comparisons from rival.tips
Source-backed AI model pricing, rankings, history, and benchmark data.
Search and read Vector Panda docs: API operations, pricing, storage tiers, measured benchmarks.
Live AI compute pricing oracle — real-time LLM model prices across providers
Audited indie product business cases with source-backed pricing, revenue and pricing benchmarks.
GPU and LLM inference benchmarks, hardware evidence, deployment recommendations, and launch configs.
Read-only MCP server for verified book recommendations and reading lists.
LLM Advisor — questions
Answers built from our own checks of this server.