Burnmeter runs on your own machine — the client starts it, so there is no endpoint to ping. Last commit 2 Nov 2026.
Local Claude Code token analytics: efficiency shares and per-model spend breakdown.
Today is the operative word: we check Burnmeter 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 burnmeter -- npx -y @reneturcios/burnmeter
{
"mcpServers": {
"burnmeter": {
"command": "npx",
"args": [
"-y",
"@reneturcios/burnmeter"
]
}
}
}
[mcp_servers.burnmeter]
command = "npx"
args = ["-y", "@reneturcios/burnmeter"]
{
"mcpServers": {
"burnmeter": {
"command": "npx",
"args": [
"-y",
"@reneturcios/burnmeter"
]
}
}
}
{
"servers": {
"burnmeter": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@reneturcios/burnmeter"
]
}
}
}
Local-first dashboard + MCP for Claude Code and Codex token usage. Per-MCP/tool breakdown.
Local MCP gateway that proxies AI clients to many MCP servers with a token-efficient code mode
Context hygiene for Claude Code: drift detection, re-read dedup, and token analytics.
Claude Code token usage, cost, and plan limits from local logs and Anthropic's usage API.
Local ledger of Claude Code and Codex tokens, API-equivalent cost, limits and token-rate changes.
Tokens and cost for Claude Code, Codex and 22 more AI coding agents, from local session files.
Self-hosted MCP task inbox shared by Claude Code, Codex, Gemini CLI and local LLMs.
Local-first memory for Claude Code and any MCP client: hybrid search + knowledge graph, $0/token.
Answers built from our own checks of this server.