Adam Prime AI
Verameter app icon

Coming soon

Verameter

A resource monitor for Apple Silicon Macs that you can talk to.

Verameter's full view. A glowing globe ringed by gauges for CPU, GPU, memory and power sits between two panels: the Mac's readings on the left, and the AI models, AI processes and busiest processes on the right. A 15-minute timeline runs along the bottom.
Verameter's full view, with sample data. Tap or click to see it full size.

Verameter shows what your Mac's processor, graphics, memory and power are doing right now, including the AI models and coding agents running on it. Ask Vera, its assistant, why your Mac is busy, or whether a model will fit before you load it, and she answers out loud. Everything runs on your Mac. Verameter only watches, explains and recommends; it never changes anything.

What it does

  • Measured power and temperatures, from the Mac's own sensors
  • Knows your local AI: Ollama, LM Studio, llama.cpp, MLX and vLLM, the models they have loaded, and whether a new one will fit
  • Names the AI coding agents at work by project, and shows which one started which job
  • Vera, a voice assistant that runs on your Mac, using Apple's on-device model or your own local model
  • Connects to AI tools such as Claude Code through its MCP server, so they can check the Mac before they act. It also offers an HTTP API and Prometheus metrics.
A close-up of Verameter's globe, ringed by gauges, with Llama 3.3 70B and Qwen3 8B loaded. Below it, Vera answers a spoken question. The question reads: Could I load a second 70B right now? The answer reads: Not beside Llama 3.3: another 70B needs about 44 GB, and 37 GB is free. A 32B model would fit.
Asking Vera whether a model will fit, with sample data.

Requirements

Requires a Mac with Apple Silicon (M1 or later) and macOS 26 or later.

Coming soon

Hear when Verameter is ready