I make Shiny, a menu-bar app for Mac memory, built with no analytics and no crash reports. Every app on this list has either a public statement, open-source code you can read, or both.
Why telemetry is the default
Most developers add analytics because product decisions are easier to defend with numbers. "We shipped this because 40% of users clicked it" is a more comfortable conversation than "it felt right." Analytics SDKs are free, take an afternoon to integrate, and produce dashboards that look authoritative.
The cost of resisting telemetry is higher: every time the team wants to know whether a feature is being used, the answer is "we don't know." That takes real discipline to maintain.
No telemetry is a small choice that has to be defended every time the team is tempted to ship one more analytics call.
For a plain-English explanation of what telemetry is and why it matters, see what is telemetry in Mac apps.
How to read this list
Each entry includes a note on how to verify the claim. The gold standard is open-source code you can read yourself. Second is a Little Snitch test: install it, launch the app, and watch what connects. A telemetry-free app makes only the connections it needs: one update check, one license ping. Everything else should be silence.
Browser: Orion by Kagi
FreeWhat it does: A native WebKit browser for macOS, built by the team behind the Kagi search engine. Supports Chrome and Firefox extensions natively, which most privacy-focused browsers don't. No built-in advertising model.
Why it's on this list: Kagi publishes an explicit no-telemetry policy for Orion. The browser does not call home with usage data. There's no analytics SDK embedded in it. Kagi's business model is subscription search, not advertising, so they have no structural incentive to collect behavioural data on how you browse.
How to verify: Run Little Snitch alongside Orion for a week of normal browsing. You'll see DNS lookups for sites you visit, and update checks to Kagi's servers, but no calls to analytics endpoints like analytics.google.com, mixpanel.com, or segment.io. The Orion source is partially available for inspection too.
Best for: anyone who wants a fully-featured browser with extension support and a clean network record. Not Firefox, not Brave, a genuine native Mac browser.
Notes: iA Writer
$49.99 onceWhat it does: A focused writing app with a clean interface, Markdown support, and very little else. One font, one layout, one job. iA Writer has been around since 2010 and the approach hasn't changed much.
Why it's on this list: iA has never shipped analytics into Writer. The app stores your files as plain Markdown on disk, no proprietary database, no sync service unless you use iCloud. There is no account to create, no dashboard to log into, and nothing calling home to iA's servers during a normal writing session. The iOS and macOS apps are distributed through the App Store, which limits what any app can access by default.
How to verify: iA Writer's network traffic, when you're not using any connected feature, is effectively zero. Run Little Snitch: the only calls you'll see are App Store receipt validation and iCloud sync if you've enabled it. Both are expected and user-initiated. There's no telemetry endpoint.
Best for: writers who want a distraction-free environment with no analytics company watching over their shoulder. A one-time purchase with no subscription.
RSS: NetNewsWire
Free, open sourceWhat it does: The oldest and most reliable Mac RSS reader. Native, fast, and syncs with multiple services including Feedbin, Feedly, and iCloud. Built and maintained by Brent Simmons, a long-standing member of the Mac indie developer community.
Why it's on this list: NetNewsWire is open source. You can read every line of it on GitHub. There is no analytics code in the repository. The app makes network calls to your chosen sync service and to the feeds you subscribe to. That's it. No ping to a NetNewsWire server, no usage data collected.
How to verify: Read the source at github.com/Ranchero-Software/NetNewsWire and search for "analytics", "telemetry", or common SDK names. You'll find nothing. Little Snitch confirms the same: the only outbound connections are your feeds and your sync service.
Best for: anyone who still reads RSS and wants the lightest, most honest RSS reader on the Mac. Free, with no catch.
PDF: Skim (or PDF Expert with analytics off)
Skim: free / PDF Expert: $79.99 onceWhat it does: Skim is a free, open-source PDF reader and annotator for macOS. It's been maintained by a small group of volunteers for over fifteen years and handles annotation, highlighting, and note-taking on PDFs well. PDF Expert is a commercial alternative with a broader feature set.
Why Skim is on this list: Skim is open source and makes no network calls during normal use. There is no telemetry code in the project. It stores everything locally. For anyone who reads and annotates academic papers or technical documents and wants a PDF reader that doesn't phone home, Skim is the honest choice.
The honest note on PDF Expert: PDF Expert, by Readdle, does collect analytics by default. You can disable this in preferences under Privacy. If you need PDF Expert's features, turn that off during setup. Skim requires no such configuration: it's clean out of the box.
Best for: researchers, students, and anyone who reads a lot of PDFs and wants a tool that stays local. Skim has no telemetry by design; PDF Expert requires you to opt out.
Memory: Shiny
$4.99 onceWhat it does: A menu-bar app that shows your Mac's memory pressure and lets you free up RAM in one click. Lives quietly in the menu bar, no background processes cluttering your app switcher. See why I built Shiny for the full story.
Disclosure: this is my app. I'm including it because it's directly relevant to the category, and because I can speak to its internals with complete confidence. Shiny has no analytics, no crash reporters, no third-party SDKs. I never see any data about how you use it.
Two visible network calls: one update check to GitHub Releases, and one license activation to the Paddle API when you first buy it. Neither sends usage data. After activation, the only ongoing call is the update check, which you can turn off in preferences.
How to verify: Run Little Snitch. You'll see exactly two endpoints: api.github.com for update checks, and checkout.paddle.com for license validation on first launch. Nothing else, the right profile for a memory utility used by thousands of people.
Best for: anyone who monitors Mac memory and wants to know their menu-bar utility isn't watching them back. $4.99 once, three Macs, no subscription.
Launcher: Alfred (or Raycast core)
Alfred: free core / Powerpack $34 once | Raycast: free coreWhat it does: Both Alfred and Raycast replace Spotlight with a faster, more capable launcher. File search, app switching, web searches, clipboard history, workflow automation, the category is mature and both apps are very good.
Alfred's position: Built by a small independent team (Running with Crayons), Alfred has had no telemetry since 2010. It doesn't connect to external servers during normal use. The business model is a one-time Powerpack purchase, so there's no subscription incentive to collect data.
The honest note on Raycast: Raycast's core launcher does not collect keystrokes or query content. However, it is VC-backed and its AI features send data to Raycast's servers when used. Stick to local launcher features and the privacy profile is reasonable. If you want no external connections at all, Alfred is the cleaner choice.
Best for: anyone replacing Spotlight and wanting a launcher they've used for years without wondering what it sends. Alfred is the quieter, older choice; Raycast is faster to set up but warrants a closer look at which features you enable.
What to look for when evaluating any app
These six categories cover the most common daily-driver apps on a Mac. For anything not on this list, the same framework applies:
Check for a public policy. Not a generic privacy policy, but a specific statement about telemetry. "We collect anonymous usage data" is telemetry. "We collect no analytics" is not.
Look for open source. Open-source apps can be verified. Search the repository for Mixpanel, Segment, Amplitude, or Firebase. If none exists, the app is almost certainly clean.
Run a network monitor for a week. Little Snitch makes every outbound connection visible and labelled by process. A week of normal use shows exactly what each app is doing.
Check the business model. Apps with advertising models or VC growth pressure have structural incentives to collect data. One-time purchase apps and small-team indie apps are less likely to benefit from telemetry.
For more on what the term means, see what is telemetry in Mac apps. For a complementary look at app choices by memory footprint, see the best Mac apps for 8GB RAM.