A native macOS performance analyzer and logger that lives in your menu bar. It continuously records CPU, memory pressure, GPU, network, disk, battery, and per-process usage to a local database, then helps you make sense of it: trends, leaks, pressure events, and on-device diagnostics.
Free and open source. No telemetry. Every sample stays on your Mac.
- Menu bar at a glance: one compact, configurable item for live memory pressure, CPU, GPU, network, disk, and battery readouts, with a shared detail panel.
- Dashboard: a plain-language verdict, headline tiles, the pressure timeline with selectable ranges, a memory breakdown, and a swap trend.
- Process explorer: a live, sortable, filterable table of every process, with a detail inspector for footprint, CPU, file descriptors, disk I/O, and Rosetta status over time.
- Process groups: group related apps and helpers into a stack and see its blended footprint as a share of the device.
- History and logging: configurable-resolution logging to a local SQLite store; top consumers over any window you pick.
- Disk tab: live throughput, IOPS, service latency, and utilization with history; per-device hardware identity; per-volume capacity bars grouped by APFS container with purgeable space; SMART health for the internal SSD; a boot-volume free-space trend; and top processes by attributed disk I/O.
- GPU tab: who is using the GPU on Apple silicon, per process, with no helper (the AGX driver's per-context accounting), plus device utilization, clock-state residency, GPU and Neural Engine power, memory, thermal limit and power cap, a breakdown by workload category, and recognition of AI runtimes (Ollama, llama.cpp, MLX, LM Studio, Core ML, Apple Intelligence) with the model they serve where the command line says; an optional sustained-high-GPU alert. The GPU menu bar dropdown lists the top GPU processes too.
- Hardware tab: this Mac's inventory as a searchable, browsable tree with a
visual overview: a block diagram of the chip (CPU clusters, GPU cores, Neural
Engine, unified memory), capacity bars, the displays to scale, battery health,
and every bus and device
system_profilerand the kernel report (USB, Thunderbolt, Bluetooth, audio, cameras, storage, network, Wi-Fi via CoreWLAN, Metal limits, instruction-set features, secure boot). Read on demand with a Refresh button, never on the sampling tick; copy any item or save a report. - Leak detection: flags processes whose footprint climbs steadily, plus a log of pressure events over time.
- Deep-dive diagnostics: explains what a process is and whether its behavior is normal, using signed, updatable check packs.
- Insights and alerts: quiet-by-default notifications for critical pressure, sustained swap, per-process ceilings, and suspected leaks.
Process explorer, with a per-process detail inspector:
Energy: battery health, an energy-flow view, and the top energy users:
Network throughput and every adapter on the machine:
Disk throughput, service latency, SMART health, free space, and top I/O processes:
GPU: utilization, power, clock states, and who is using the GPU, with AI workloads picked out:
Hardware: this Mac's inventory, searchable, with the chip drawn core by core:
Analytics: build your own per-process charts over any window:
Insights: what changed, pressure events, and the heaviest consumers:
Download MacPerformanceMonitor.pkg from the Releases page and
double-click it. It's Developer ID signed and notarized by Apple, so it installs and
launches without security warnings, and keeps itself up to date via Sparkle.
brew tap zesty0wl/mac-performance-monitor https://github.com/Zesty0wl/mac-performance-monitor
brew install --cask zesty0wl/mac-performance-monitor/mac-performance-monitorThis installs the same signed, notarized pkg. A submission to the main
homebrew-cask repository is in
progress (see docs/homebrew-submission.md); once accepted, the tap step won't
be needed and brew install --cask mac-performance-monitor will just work.
git clone https://github.com/Zesty0wl/mac-performance-monitor.git
cd mac-performance-monitor
swift build
swift test
Scripts/run.shRequires macOS 15 (Sequoia) or later and a Swift 6 toolchain (Xcode 16 or a Swift.org toolchain), on Apple silicon.
No telemetry, no analytics, no phone-home. Every sample is written to a local SQLite database and never leaves your Mac. Being open source, anyone can audit exactly what it does.
Contributions are welcome. See CONTRIBUTING.md and the Code of Conduct. Security reports go through SECURITY.md.
Released under the MIT License. Bundles GRDB.swift (MIT) and Sparkle (MIT).








