TokenTelemetry is an open-source developer tool designed to give you deep visibility into what your AI coding agents think, cost, and do. Instead of requiring complex setups, TokenTelemetry works by reading the session logs that your AI coding agents already write directly to your disk, transforming them into a comprehensive local dashboard.
With support for 13+ popular agents—including Claude Code, Codex, Cursor, Copilot, Gemini CLI, Antigravity, Grok, OpenCode, Cline, Qwen, Pi, and Hermes—the tool tracks crucial metrics such as token usage, cost accumulation, step-by-step trace replays, subagents, skills, and Model Context Protocol (MCP) calls.
Built with developer privacy and simplicity in mind, TokenTelemetry requires no SDK, no code changes, and no cloud account. It can be installed with a single command, keeping all of your data entirely secure and local on your machine.
TokenTelemetry is an open-source local dashboard that reads session logs written to disk by AI coding agents to track metrics like token usage, costs, subagents, MCP calls, and step-by-step traces.
It supports 13+ agents including Claude Code, Codex, Cursor, Copilot, Gemini CLI, Antigravity, Grok, OpenCode, Cline, Qwen, Pi, and Hermes.
No. TokenTelemetry requires no SDK, no code changes, and no cloud account. Everything runs locally on your machine and can be installed with a single command.
0 community reviews for TokenTelemetry
Help the community by sharing your experience. Your review helps others make better decisions.
On a mission to make learning faster
