Clawdmeter turns Claude Code token stats into a DIY desktop dashboard
A new open-source gadget built on the ESP32-S3 displays real-time Claude Code usage on a 2.16-inch AMOLED screen, complete with an animated pixel-art mascot.
What matters
- Developer Herman Haraldsson released Clawdmeter as an open-source project on GitHub.
- It uses a Waveshare ESP32-S3-Touch-AMOLED-2.16 module with a 2.16-inch, 480×480 AMOLED display.
- The device tracks Claude Code token usage via Anthropic API response headers and displays it with an animated pixel mascot named Clawd.
- It connects over Bluetooth Low Energy and doubles as a BLE HID keyboard for shortcuts.
- The project reflects the growing “tokenmaxxing” trend among developers.
Web developer Herman Haraldsson has unveiled Clawdmeter, an open-source desk gadget that monitors Claude Code token usage in real time. The device is built around the Waveshare ESP32-S3-Touch-AMOLED-2.16, a 2.16-inch AMOLED touchscreen module powered by an Espressif ESP32-S3R8 dual-core microcontroller clocked at up to 240 MHz. According to project details shared on GitHub and reported by CNX Software and dev.ua, the board includes 512 KB of RAM, 8 MB of PSRAM, 16 MB of NOR flash, and a microSD card slot, with connectivity handled over 2.4 GHz Wi-Fi 4 and Bluetooth 5.0 LE.
The dashboard pairs with a laptop over Bluetooth Low Energy and functions as a BLE HID keyboard for shortcuts. It pulls limit data directly from Anthropic API response headers and renders statistics through the LVGL graphics library. A pixel-art mascot named Clawd appears on the 480×480 screen, changing its behavior based on how actively the programmer is using Claude Code. Physical buttons on the unit can trigger quick commands, such as switching modes in Claude Code voice mode with Space and Shift+Tab combinations.
What happened
Clawdmeter is more than a novelty. It is a physical artifact of how deeply Anthropic’s Claude has embedded itself in developer workflows, and of the emerging “tokenmaxxing” culture—where software engineers increasingly treat high token consumption as a proxy for productivity and AI fluency. As TechCrunch notes, terminal commands and external apps can already track usage, but a dedicated hardware dashboard signals that AI metering has become important enough to earn a permanent spot on the desk.
The project also highlights the maturity of off-the-shelf ESP32 hardware ecosystems. Because Clawdmeter relies on standard components and open-source firmware, it lowers the barrier for developers to build custom AI telemetry tools without fabricating custom PCBs. The software has been fully tested on Linux (Ubuntu), with macOS support currently under development, which suggests a small but growing audience of power users wants ambient, always-visible visibility into API spend.
Why it matters
No strong public signal was available from Reddit or broad forum discussions at the time of publication. TechCrunch cited one Reddit user joking that Anthropic should mail the devices to developers for free, and another suggesting—only half in jest—that adding a button to top up tokens with a card on file could be dangerous.
Public reaction
Watch whether Anthropic formalizes hardware partnerships or releases an official SDK for physical telemetry devices. The company has not endorsed Clawdmeter, and the project’s reliance on API response headers means any change to Anthropic’s rate-limit headers could break compatibility without warning. Also monitor macOS support progress, since the current Linux-only testing limits the gadget’s addressable market among developers. Finally, observe whether “tokenmaxxing” remains a playful subculture or evolves into a formal productivity metric inside engineering organizations, which could pressure vendors to build native usage widgets rather than leave the ecosystem to DIY solutions.
What to watch
Sources
Public reaction
Public discussion remains light, with only scattered social media reactions captured. A Reddit user quoted by TechCrunch joked that Anthropic should distribute the hardware for free, while another quipped about adding a one-click token top-up button. No broad developer forum consensus or criticism has emerged yet.
Signals
- Mild amusement at the hardware form factor
- Curiosity about Bluetooth and HID keyboard integration
- Playful concern over spend tracking and token top-ups
Open questions
- Will Anthropic support or tolerate third-party hardware polling its API headers?
- How accurate is the token counting compared to Anthropic’s own dashboards?
What to do next
Developers
Build or fork the Clawdmeter firmware to adapt it for other LLM APIs that expose usage headers, using the existing ESP32-S3 and LVGL stack.
The open-source firmware and standard hardware make it straightforward to repurpose for OpenAI, Gemini, or internal API metering.
Founders
Evaluate whether your AI tool’s API exposes real-time usage headers; transparent metering can inspire community hardware projects that reinforce brand loyalty.
When power users build physical dashboards for your API, it signals deep workflow integration and creates organic evangelism.
PMs
Study Clawdmeter as an example of user-led product extension—identify which power-user metrics are important enough that customers build hardware to track them.
Shadow dashboards reveal unmet needs in native analytics and can guide roadmap priorities for official usage tools.
Investors
Treat the “tokenmaxxing” trend and community hardware projects as early signals of platform stickiness for Anthropic’s developer ecosystem.
Developer willingness to dedicate desk space to usage tracking suggests high engagement and switching costs.
Operators
Audit whether your engineering teams are relying on informal token metrics; consider formalizing cost and usage visibility before shadow dashboards proliferate.
DIY telemetry often emerges when official tooling is insufficient, and unmonitored API spend can surprise finance teams.
How to test
- 1Clone the Clawdmeter repository from GitHub.
- 2Install dependencies for LVGL and NimBLE.
- 3Configure Wi-Fi and Bluetooth credentials in the firmware.
- 4Build and flash the firmware to the ESP32-S3 board.
- 5Pair the device with your computer via BLE.
- 6Run Claude Code and authorize API header polling.
- 7Observe token usage and Clawd animations on the AMOLED screen.
Caveats
- macOS support is not yet fully tested
- The device depends on Anthropic API response headers that could change without notice
- Hardware must remain within Bluetooth range of the host computer
- Pixel-art animations are cosmetic and do not affect core functionality