Clawdmeter turns Claude Code usage stats into a desktop dashboard

Clawdmeter turns Claude Code from a background assistant into something developers can watch like a gauge on a dashboard. The small open source device, built by Hermann Bjorgvin, uses a Waveshare ESP32-S3-Touch-AMOLED-2.16 display to show token usage on a tiny desktop screen, with pixel-art Clawd animations that become more active as the meter rises.

That visual readout matters because Claude Code has become more than a chat window. Anthropic describes it as an agentic coding system that can analyze codebases, make multi-file changes, run tests and complete development tasks autonomously. As more engineers hand off larger chunks of work to the tool, the questions shift from novelty to accountability: how much is it using, how fast is it burning through limits, and what does that dependence cost?

Clawdmeter answers those questions by pairing with a laptop over Bluetooth Low Energy and reading limit-usage data directly from Anthropic API response headers. Its two side buttons can send Space and Shift+Tab over BLE HID, giving users hardware shortcuts for Claude Code voice mode and mode switching. Adafruit said the device monitors token usage on a 2.16-inch AMOLED screen and uses the LVGL graphics library and the NimBLE stack, while CNX Software described it as a DIY firmware project built on off-the-shelf Waveshare hardware.

The project lands in the middle of a broader wave of third-party monitoring tools built around Claude Code. Local dashboards and widgets are tracking tokens, costs, sessions and rate limits because Anthropic’s own usage analytics are still limited. According to Anthropic’s help-center guidance, those analytics are available to Team plans, Enterprise owners and admins, and API Console roles, but not to individual Pro or Max plans.

That gap helps explain the appetite for gadgets like Clawdmeter. Anthropic said on May 14, 2026 that it was doubling Claude Code’s five-hour rate limits for Pro, Max, Team and seat-based Enterprise plans and removing peak-hours limit reductions for Pro and Max accounts. For power users, that makes usage visibility more than a convenience. It becomes a daily signal about whether AI coding is speeding up work, consuming budget, or bumping up against the boundaries of the subscription itself. The repository had already attracted attention, with OSSInsight listing about 699 stars and 42 forks, a sign that the market for measuring AI work is now part of the work itself.