Does Slack drain your Mac's battery?

Measured 12–14 and 21–23 August 2026 · two separate runs on the same Apple silicon MacBook · 326 windows with Slack running · reproducible with the script at the bottom

Short answer: almost never, and the exception is not dependable either. Slack has a reputation as a battery hog, and its measured battery usage on a Mac says the reputation is built on the minutes it deserves rather than the hours it doesn't. We measured the same machine twice with WattMate, nine days apart: the typical figure came back identical, and the alarming one did not come back at all.

Half the time Slack cost less than a hundredth of a watt. On a 60 Wh battery that is, quite literally, days of runtime before it adds up to one percent. That figure is the sturdy one: two runs, nine days and 496 windows apart, both put the median at 0.01 W.

The spike is the fragile one. In August's first run a handful of windows reached 7.4 W, more than the entire rest of that Mac was using at idle: a 736-fold gap between typical and worst. In the second run the worst window was 1.2 W, and the whole period cost 0.16 Wh against 1.1 Wh before. Per window that is eight times cheaper, and the difference is almost entirely those few spikes: the median never moved.

So which one is "the truth"?

Both, and that's the point. An app with a spiky profile is invisible in averages and unmistakable in the moment. If you looked at Slack during one of those spikes you'd conclude it was eating your battery; if you looked at any of the other 150 windows you'd conclude it does nothing. Neither snapshot is wrong; both are useless on their own.

What we deliberately can't tell you is why the spikes happened. The log records what each app used, not what you were doing. So "it was a huddle" would be a guess, and a plausible guess is still a guess. What we can say is that they were brief and rare, which is why the total across the whole period is barely over one watt-hour.

Why macOS says "Slack is using significant energy" and then stops

The battery menu names an app while its Energy Impact stays high, and Energy Impact is a relative score with no unit: there is no published threshold to check a reading against. On an app shaped like Slack that produces a label which appears and vanishes without anything meaningful having changed.

The spikes above are the whole explanation. During one of them Slack sits at the top of that list; across the other 150 windows it does nothing at all. The menu reports the moment you happened to open it, and the total across the whole period is barely over one watt-hour.

What this means for your afternoon

If your battery is disappearing and Slack is open, Slack is probably not the reason, and quitting it will most likely return you nothing, while costing you the messages. On the same machine, in both runs, Chrome was using around four watts as a matter of routine, four hundred times Slack's median: that page has the numbers.

The useful move isn't to quit the app with the worst reputation. It's to look at what your Mac is using right now, in watts, as WattMate lists it, and quit whatever is actually at the top, which on most machines is a browser, an editor or a video call, not a chat client sitting in the background.

What this does and doesn't prove

Still one Mac, one workspace. Two runs are not two machines. Slack is installed on only one of the Macs behind these pages, so unlike our Chrome page there is still no second machine to disagree with this one. What the repeat buys is narrower: it says the median is a property of the app rather than of one particular week. A Mac in busier workspaces, with more calls or more open channels, would spike more often.

The spike did not reproduce, and we are not deleting it. The honest reading of two runs is that the near-zero baseline is reliable and the 7.4 W window is an event, not a rate: it happened, we cannot say why, and it did not happen again in the next 339 windows. Treat the shape (near-zero with occasional spikes) as the finding, and either set of exact figures as one example of it.

The spikes are real and we are not smoothing them away. It would be easy to publish just the median and call Slack free; the 7.4 W window happened, and an app that occasionally costs that much on battery deserves to be seen when it does. It is equally easy to publish only the scary number and never check whether it returns. We did check. That is what a live measurement is for.

Measure your own

Every figure here comes from one script over WattMate's own database:

python3 Scripts/app-energy-report.py <database> Slack

Or watch it live: the trial ranks every app by actual watts and tells you what quitting the top one is worth, then checks its own answer two minutes later.

One thing to expect when you do: WattMate's number for the same program is lower than the table above, and neither reading is wrong. This page measures at the chip's own power rails. The app reports the share of the battery's drain it can put on that program once the floor the Mac pays anyway is taken out, and that is the figure minutes are made of. The ranking is the same either way. The scale is not.

Related: what Telegram costs, thirty times this · what Chrome costs · what Spotify costs · does quitting apps help at all? · what every app costs, measured