A terminal-based system monitor for Apple Silicon Macs running Ollama. Tracks RAM, swap, CPU, SSD usage, and installed Ollama models in a live-updating dashboard.
+----------------------------+------------------------------------------+
| π§ Memory [12:34] | βοΈ CPU |
| RAM Used 54.2 / 64 GB | Overall [ββββββββββββββββββββ] 42.0% |
| RAM Free 9.8 GB | Core 0 [ββββββββββββββββββββ] 31.2% |
| Swap Used 0.3 / 4.0 GB | Core 1 [ββββββββββββββββββββ] 55.8% |
| | Core 2 [ββββββββββββββββββββ] 18.3% |
+----------------------------+------------------------------------------+
| πΎ Storage | π¦ Ollama β Installed Models |
| SSD Used 412 / 994 GB | llama3:70b 40 GB β RUNNING |
| SSD Free 582 GB | qwen2:32b 20 GB β idle |
| Total Read 18204 MB | mistral:7b 4 GB β idle |
| Total Write 12840 MB | |
+----------------------------+------------------------------------------+
| π Model RAM Guide (Q4, Apple Silicon) |
| 7B .... 4.5 GB β
fits 64GB |
| 13B ... 8 GB β
fits 64GB |
| 70B ... 40 GB β
fits 64GB |
| 180B .. 110 GB β no |
+-----------------------------------------------------------------------+- Real-time RAM usage via
vm_stat(matches Activity Monitor, not justpsutil) - Per-core CPU utilization
- SSD usage and cumulative I/O stats
- Swap memory monitoring
- Lists all installed Ollama models with running/idle status
- Q4 quantization RAM guide for model sizing
- Color-coded progress bars (green β yellow β red)
- Refreshes every 2 seconds
- macOS on Apple Silicon
- Python 3.10+
- Ollama installed (optional β the monitor works without it)
pip install psutil rich./sysmon.pyOr:
python3 sysmon.pyPress Ctrl+C to exit.
| Metric | Source |
|---|---|
| RAM | vm_stat + psutil (wired + active + compressor pages) |
| CPU | psutil.cpu_percent(percpu=True) |
| Disk | psutil.disk_usage + psutil.disk_io_counters |
| Swap | psutil.swap_memory |
| Ollama models | ollama list + ollama ps |
RAM is calculated from vm_stat rather than relying solely on psutil to match what Activity Monitor reports on macOS.
MIT