NythyCleaner
← Blog

How to Optimize Your Mac in 2026 — Complete Guide

18 min read
#optimize-mac#mac-performance#mac-cleanup#macos#monitoring#privacy#xcode#tutorial

How to Optimize Your Mac in 2026

Want to optimize your Mac instead of just deleting a few caches? Real macOS optimization means reclaiming storage, trimming developer leftovers, controlling startup load, checking privacy traces, and monitoring CPU, memory, GPU, network, and disk activity in one place.

This guide is a full walkthrough of how NythyCleaner helps optimize a Mac in 2026, from system cleanup and Xcode cleanup to disk treemaps, iPhone storage scans, monitoring, maintenance, privacy, and uninstallation. If your immediate goal is storage, start with how to free up disk space on Mac or how to free up iCloud Drive space on Mac.


1. System Cleanup — Deep, Category-by-Category Cleaning

The System Cleanup section does not treat your Mac as a black box. It breaks down clutter into 14 distinct categories, each with a clear description and independent toggle:

  • User caches — per-app caches under ~/Library/Caches (Safari, Chrome, Spotify, Xcode, and hundreds more).
  • User logs — diagnostic logs from apps and the system stored under ~/Library/Logs.
  • User temporary files — leftover temp data created by apps.
  • Downloads — forgotten .dmg installers, ZIP archives, and attachments piling up in ~/Downloads.
  • Trash — the Finder Trash, which still occupies disk space until emptied.
  • System temporary files/private/tmp and other transient system data.
  • Shared caches — system-level caches in /Library/Caches (may require admin).
  • System logs — logs under /Library/Logs and /private/var/log.
  • Browser caches — cached media, images, and scripts from Chrome, Firefox, and Safari.
  • Browser data — cookies, local storage, history databases, and session data.
  • Old iOS backups — local device backups that can consume tens of gigabytes.
  • Mail attachments — downloaded attachments stored by Apple Mail.
  • VM files — swap files and sleep images under /private/var/vm.
  • Safari extensions and plugins — outdated or unused browser extensions.

Beyond simple deletion

The scan also runs two secondary passes:

  1. Largest files scanner — surfaces the biggest individual files across user directories so you can spot space hogs that do not belong to any cache category.
  2. Duplicate-size groups — files of identical byte size in ~/Downloads and Mail paths, flagged as potential duplicates for review (fast pre-filter, no hashing overhead).

You can exclude specific paths from scanning permanently, and the app tracks a cleanup history with a chart so you can see how much space you have reclaimed over time.


2. Xcode & Developer Tools — Three Cleanup Engines in One Screen

If you write code on a Mac, developer tools quietly consume enormous amounts of storage. NythyCleaner dedicates an entire section to this, split into three tabs.

Xcode cleanup (22 categories)

Every known Xcode data folder is scanned and measured:

  • DerivedData — the single biggest offender, rebuild caches for every project.
  • Archives — old .xcarchive builds you no longer distribute.
  • Device support — symbols for every iOS, watchOS, tvOS, and visionOS version you have ever connected.
  • Simulator caches — CoreSimulator runtime data, cached assets, and old simulator runtimes that Apple never removes automatically.
  • Xcode application caches, Instruments caches, IDE editor cached data, SwiftUI previews cache.
  • SwiftPM caches (local and global), Carthage, CocoaPods, Tuist, Mint, Fastlane.
  • Documentation cache, device logs, Xcode UserData backups, provisioning profiles.
  • Old simulator runtimes — detected via simctl and removable individually.

Homebrew cleanup

The built-in Homebrew tab performs a dry-run analysis, shows the formula cache size, and lets you run brew cleanup with full visibility into what will be removed — including outdated packages and stale downloads.

Docker cleanup

If Docker Desktop is installed, the app detects whether the daemon is running, measures disk usage via docker system df, and offers operations to remove dangling images, stopped containers, and unused volumes.


3. Disk Space — Native Treemap, Volume Actions, and iPhone Scanning

This is where NythyCleaner diverges from anything else on the market.

Interactive treemap (C++ squarify engine)

The disk visualization is not a chart library running in JavaScript. The layout is computed by a native C++ squarify algorithm (NativeTreemap.cpp) that produces the tile positions for every folder and file on the volume. The result is an interactive, zoomable treemap rendered in SwiftUI — each colored block proportional to the folder's size.

You can drill down into any directory using a breadcrumb bar. Large clusters of space become immediately visible without reading a single file listing.

Volume management

NythyCleaner reads every mounted volume and provides direct actions:

  • Mount / unmount / eject external drives.
  • Rename a volume.
  • View detailed volume info — file system type, capacity, free space, mount flags.
  • Verify and repair a volume using diskutil (admin privileges granted through the privileged helper).
  • Time Machine snapshots — list and manage APFS snapshots occupying hidden space.

iPhone and iPad storage scanning

Plug in an iOS device via USB, and NythyCleaner uses a native C++ MobileDevice / AFC bridge (IOSDeviceManager.cpp) to:

  • List connected devices with model and capacity info.
  • Enumerate every file on the device and feed the data into the same treemap engine — so you get the same interactive visualization for your iPhone as you do for your Mac's SSD.

This is not a simple "used vs. free" bar — it is a full file tree walk over AFC.

Expert Disk Utility

For power users, a dedicated expert sheet exposes the full topology of your disks (containers, volumes, partitions) and lets you build diskutil commands interactively:

  • Erase a volume or disk.
  • Partition a disk.
  • APFS operations — add/delete/resize APFS volumes and containers.
  • Mount / unmount with options (read-only, no-browse).
  • Verify and repair at the container or volume level.

Every command is validated before execution, and APFS-specific operations use the correct diskutil apfs subcommands. Privileged operations are routed through the app's signed helper.

Format utility

A simplified sheet for formatting a volume: choose the file system (APFS, HFS+, ExFAT, FAT32), name, and scheme — with guard rails to prevent accidental data loss on the boot volume.

Interesting files scanner

A secondary scan surfaces "interesting" files on the volume — unusually large media, old archives, orphaned app bundles, and other candidates for review — so you do not have to browse every folder manually.


4. Real-Time Monitoring — 10 Tabs, Hardware-Level Telemetry

The Monitoring section is a full system observatory with 10 specialized tabs.

Overview

A dashboard combining CPU load, memory pressure, disk I/O, network throughput, and temperature at a glance.

CPU

  • Total usage computed from Mach host tick deltas (host_statistics).
  • Per-core usage — individual load for every efficiency and performance core.
  • Load average (1 / 5 / 15 min).
  • Process and thread count.
  • CPU core control — on supported hardware, NythyCleaner can take individual cores offline or bring them back online using cpuctl through the privileged helper. This is useful for thermal management, power saving, or benchmarking.

Memory

  • Physical vs. used vs. wired vs. compressed breakdown.
  • Swap usage and pressure indicators.

GPU

  • GPU utilization metrics read from IOKit (IOAccelerator).

Neural Engine (ANE)

  • ANE utilization when available — useful for monitoring Core ML workloads.

Network

  • Per-interface throughput (bytes in/out, packets, errors) with delta computation.

Disk

  • I/O metrics and volume health data from IOKit and SMART attributes.

Processes

  • Live process list with CPU and memory usage per process.

Battery

  • Cycle count, health percentage, charging state, and power source details via BatteryMetricsCollector.

History

  • All metrics are persisted (MonitoringHistoryStore) and plotted over time, so you can spot trends — a CPU that runs hot every afternoon, memory pressure that grows over days, or a disk that is slowly filling.

Alerts

  • Configurable thresholds (MonitoringAlertPreferences) trigger alerts when CPU, memory, disk, or temperature exceed your chosen limits.

Menu bar integration

The menu bar extra displays a live CPU gauge in the system tray. Clicking it shows compact cards for CPU, RAM, and boot disk usage, plus the date of your last cleanup — with a quick-scan shortcut that jumps straight to System Cleanup.


5. Privacy Audit — Permissions, Traces, and Secure Deletion

The Privacy section is not just "clear browser history." It performs a structured audit across four dimensions:

  1. TCC permissions — reads the macOS Transparency, Consent, and Control database to show which apps have access to your camera, microphone, screen recording, Full Disk Access, contacts, calendar, and more.
  2. System traces — surfaces recent files lists (SharedFileList plists), shell command history (~/.zsh_history, ~/.bash_history), Spotlight recent searches, Quick Look thumbnails, and clipboard contents.
  3. Network history — known Wi-Fi networks, DNS cache, and related network artifacts.
  4. Browser data — cookies, local storage, autofill, and history across Safari, Chrome, and Firefox.

Each dimension contributes to a privacy score visualized with charts. You can selectively clean any subset, and sensitive files are removed using secure deletion (SecureDeleteService) that overwrites data before unlinking.


6. Duplicate Finder — Vision AI Similarity, Not Just Hash Matching

Most duplicate finders compare files byte-for-byte or by hash. NythyCleaner uses Apple's Vision framework to compute feature prints (perceptual embeddings) for images and videos, then clusters visually similar media using SIMD-accelerated Euclidean distance computations in C++ (FeaturePrintDistance.cpp).

This means it catches:

  • Photos resized to different dimensions.
  • Screenshots cropped differently.
  • Images re-encoded in another format (JPEG vs. HEIC vs. PNG).
  • Videos with different compression but identical content.

The underlying file comparison engine also supports parallel byte-by-byte comparison and SHA-256 hashing (native C++) for exact-match scenarios — but the Vision-based similarity search is the differentiator.


7. Uninstaller — Full Residual File Detection

Dragging an app to the Trash leaves behind preferences, caches, Application Support folders, containers, saved state, and login items. The Uninstaller scans all known residual locations for a selected app:

  • ~/Library/Application Support/
  • ~/Library/Caches/
  • ~/Library/Preferences/ (including ByHost)
  • ~/Library/Saved Application State/
  • ~/Library/Containers/
  • ~/Library/Group Containers/
  • Login items and LaunchAgents referencing the app.

You choose which residual groups to remove — the app never deletes silently.


8. Extensions & Startup Management

The "Extensions" section goes beyond simple login items:

  • Login items — apps and helpers that launch at login (with the ability to disable/re-enable).
  • LaunchAgents — user-level and system-level plists in ~/Library/LaunchAgents and /Library/LaunchAgents.
  • LaunchDaemons — system daemons in /Library/LaunchDaemons.
  • Quick Look plugins — generators registered for file previews.
  • Spotlight plugins — importers that index custom file types.

Invalid or orphaned entries (broken plist, missing executable) are detected and can be purged in one action.


9. System Maintenance — 50+ One-Click Tasks

The Maintenance section offers a curated library of macOS administration tasks, each explained and executable with a single click. Here is a sample:

System repair and optimization

  • Flush DNS cache
  • Rebuild LaunchServices database
  • Reindex Spotlight
  • Purge inactive memory
  • Purge icon caches, font caches, Quick Look cache
  • Repair home folder permissions
  • Rebuild Mail database
  • Rebuild DYLD shared cache
  • Verify disk volume
  • Reset SMC / NVRAM
  • Thin Time Machine snapshots

Network

  • Reset Wi-Fi / Ethernet
  • Clear ARP cache
  • Flush routing table
  • Renew DHCP lease
  • Restart Bluetooth

Developer tools

  • Clear npm / pip / Yarn caches
  • Clear Terminal history
  • Remove .DS_Store files recursively

Finder & Desktop

  • Reset Dock / Launchpad
  • Show hidden files / file extensions / Library folder / Path Bar / Status Bar
  • Keep folders on top
  • Disable desktop icons
  • Reset Finder preferences
  • Clear recent items and recent folders

Media & peripherals

  • Restart CoreAudio
  • Restart Camera
  • Clear print queue
  • Change screenshot format, location, or shadow

Security

  • Disable / re-enable Gatekeeper
  • Clear clipboard
  • Disable Notification Center

App updates (second tab)

  • Aggregates outdated apps from Homebrew, Sparkle update feeds, Electron/GitHub releases, and the Mac App Store — plus dev package security audits.

10. Scheduled Cleanup — Set It and Forget It

Select categories to clean automatically — user caches, user logs, /private/tmp, DerivedData, Xcode archives, Homebrew cache, Trash — and NythyCleaner handles the rest.

A LaunchAgent (io.nythycleaner.scheduled-cleanup) wakes the app every four hours in the background. The app checks whether a run is due based on your chosen interval, executes the selected cleanup tasks silently, and logs results. No manual intervention needed after initial setup.

💡

Scheduled Cleanup is a Pro feature. The LaunchAgent is automatically installed or removed based on your subscription status.


11. iCloud Drive Cleanup

A dedicated section scans the local iCloud Drive root (~/Library/Mobile Documents/com~apple~CloudDocs), surfaces files by size and type, and lets you:

  • Move to Trash — standard safe deletion.
  • Delete — when the file location allows it.
  • Evict the local copy — keep the file in iCloud but free the disk space on your Mac.

This bridges the gap between the macOS Storage panel (which shows quota) and actual local disk pressure.


12. System Information

Hardware and software details in one place:

  • CPU — model, core count (performance + efficiency), architecture.
  • Storage — volumes, capacity, file system.
  • USB — connected peripherals.
  • Bluetooth — paired devices.
  • Startup disk — boot volume details.
  • General macOS version and build information.

13. Under the Hood — What Makes It Different

Native C++ core

Performance-critical operations are implemented in C++, exposed to Swift through a bridging layer:

ModulePurpose
NativeTreemap.cppSquarify treemap layout algorithm
DirectoryTreeWalk.cppRecursive disk usage measurement
DevDependencyWalk.cppDeveloper dependency tree analysis
FileHashing.cppSHA-256 hashing (optimized)
FileComparison.cppByte-by-byte file comparison (parallel)
FeaturePrintDistance.cppSIMD Euclidean distance for Vision feature prints
IOSDeviceManager.cppMobileDevice + AFC for iPhone file enumeration

Privileged helper

Operations that require root — disk repair, volume formatting, CPU core control, certain maintenance tasks — are routed through a signed, SMJobBless-installed helper (NythyPrivilegedHelper). The helper enforces a strict allowlist of commands and subcommands, so it cannot be repurposed as a general shell escalation.

Full Disk Access aware

The app detects Full Disk Access status and guides you through granting it via an onboarding flow. Many scan categories produce deeper results with FDA enabled.

Localized in 15 languages

Every label, description, and alert is localized via Localizable.xcstrings — English, French, German, Spanish, Italian, Portuguese, Dutch, Polish, Japanese, Korean, Russian, Traditional Chinese, Simplified Chinese, Swedish, and Norwegian.


Conclusion

NythyCleaner is not a single-purpose utility. It is a system-wide optimization toolkit that combines:

  • Deep, category-aware cleanup for system files, developer tools, and browsers.
  • A native-performance disk visualizer with iPhone scanning.
  • Real-time hardware monitoring with per-core control.
  • A privacy audit that reads TCC permissions and surfaces hidden traces.
  • Vision-powered duplicate detection that finds similar media, not just identical files.
  • 50+ maintenance tasks that would otherwise require Terminal commands.
  • Automated scheduled cleanup that runs silently in the background.

Whether you are a developer reclaiming 40 GB of DerivedData, a photographer hunting near-duplicate images, or a power user who wants to control which CPU cores are active — the answer is in one app.