SiliconLM

Not currently deployed · Source retained

Power at the Edge.

The native dashboard for Local LLM operations on Apple Silicon.
Machine status, model management, Ollama visibility, and OpenCode profiles in one place.

View Documentation View Source

Machine Info

Chip model, GPU cores, Neural Engine, RAM, and disk usage at a glance. Optimized for M-series chips.

Ollama Status

Read-only visibility into the local Ollama API when it is running. SiliconLM never starts or stops the backend.

Dashboard Only

Dashboard-only by design: SiliconLM does not proxy inference or own a runtime backend.

Status & Tools

Monitor Ollama and manage OpenCode profile/service visibility from the dashboard.

No Runtime Proxy

SiliconLM deliberately avoids owning chat proxy traffic. Use your runtime backend directly for inference.

Model Downloads

Search HuggingFace and download models with parallel acceleration. Resume interrupted downloads.

Ready to run?

terminal — -zsh — 80x24

$ git clone https://github.com/nxxxsooo/siliconlm.git

$ cd siliconlm

$ ./install.sh

Requires macOS 14.0+ and Apple Silicon (M series)

For AI Agents

Copy the prompt below and paste it into your AI assistant. It will handle the rest.

ai-agent-prompt

Install and set up SiliconLM on my Mac.

Repository: https://github.com/nxxxsooo/siliconlm

Steps:
1. Clone the repo to ~/Documents/sync/GitHub/siliconlm (or ask me where)
2. Create a Python venv and install requirements.txt
3. Start the dashboard (server.py on port 1234)
4. Confirm the dashboard starts and shows OpenCode plus optional Ollama read-only status

Requirements:
- macOS 14.0+ with Apple Silicon (M series)
- Python 3.10+
- No API keys or secrets needed

After setup, open http://localhost:1234 to verify the dashboard is running.

Shell alias to add:
alias slm='cd ~/Documents/sync/GitHub/siliconlm && nohup .venv/bin/python server.py > /tmp/siliconlm.log 2>&1 & sleep 2 && open http://localhost:1234'