Modelcostsaver runs on your own machine — the client starts it, so there is no endpoint to ping. 58 installs a week from npm. Last commit 23 Jun 2026.
Offline MCP: predict LLM call cost and pick the cheapest capable model. No keys, no network.
Nothing serious here today
Today is the operative word: we check Modelcostsaver 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 modelcostsaver -- npx -y @workswarm/modelcostsaver{
"mcpServers": {
"modelcostsaver": {
"args": [
"-y",
"@workswarm/modelcostsaver"
],
"command": "npx"
}
}
}[mcp_servers.modelcostsaver]
command = "npx"
args = ["-y", "@workswarm/modelcostsaver"]{
"mcpServers": {
"modelcostsaver": {
"args": [
"-y",
"@workswarm/modelcostsaver"
],
"command": "npx"
}
}
}{
"mcpServers": {
"modelcostsaver": {
"args": [
"-y",
"@workswarm/modelcostsaver"
],
"command": "npx"
}
}
}Alternatives to Modelcostsaver
same job, measured the same wayCut LLM token costs: count tokens, estimate cost, slim prompts, and pick the cheapest capable model.
Local LLM cost proxy: forecasts each call and recommends the cheapest model that clears the task.
Sourced AI-model pricing and capability data — compare and route to the cheapest capable model.
Live LLM API pricing: token prices, comparisons, cheapest-model lookups. No key required.
LLM cost optimization. Audit your calls, prove savings, route to the cheapest quality model. BYOK.
LLM cost optimization. Audit your calls, prove savings, route to the cheapest quality model. BYOK.
Connect Claude to any OpenAI-compatible LLM endpoint and offload routine work to a local model.
Durable agentic memory, encrypted at rest. Fully offline: no network, no API key, no cloud.
Modelcostsaver — questions
Answers built from our own checks of this server.