Is the MacBook Air M2 throttling a defect?

No, it is the cost of a silent, fanless, thin machine, paid deliberately. Every fanless computer makes this trade. It only becomes a problem when the buyer's workload is sustained-heavy, which is what the Pro line with fans exists for.

Do thermal pads fix M2 Air throttling?

The aftermarket pad mod measurably improves sustained performance by coupling the board to the case better, and it voids expectations of warranty, raises case temperature on your actual lap, and turns the chassis into more of the heatsink than Apple designed. It works; whether it is wise is a different question. We note it exists rather than recommend it.

M2 Air vs M2 Pro for video editing?

For short clips, the Air is genuinely fine, the work bursts. For long exports, the Pro's fan means it holds its level while the Air settles down a step. The deciding question is your timeline length, and a sustained measurement of an Air answers exactly how much the difference would be for you.

Does the M1 or M3 Air have the same issue?

The whole fanless Air line shares the physics; each generation moves the numbers, not the shape. Chip-specific figures deserve chip-specific pages; the measured method for finding YOUR unit's settle point is identical on all of them.

Problems · Specific machines

Why is my MacBook Air M2 slow?

An M2 Air that feels slow under real work is usually meeting its one design limit: it has no fan. Under sustained all-core load it begins stepping down within about five minutes and can settle roughly 25% below its peak: documented behaviour, not a defect in your unit.

For bursty work, browsing, documents, short exports, the M2 Air stays brilliant, because bursts finish before heat arrives. The slowness lives specifically in long sustained jobs: big exports, long compiles, extended gaming.

First, notice when it happens

From the first minute

Slow at everything from the start is NOT the fanless story, check storage pressure (a nearly-full SSD), memory pressure in Activity Monitor (8GB models swapping hard), and what launch agents run at login.

Only after minutes of use

Fast then settling lower during long heavy work: the fanless design speaking. The chassis is the heatsink; once it is warm, the chip steps down and holds a sustainable level until the load ends.

The causes, ranked

1. Design throttling under sustained load

most common

How to confirm it

The machine is fine for minutes then slows during long exports or games; the chassis is warm above the keyboard; it recovers after rest. Reviewers measured the same curve on every unit: up to ~25% down under prolonged load.

The fix

No repair fixes a design. What helps at the margin: cool room, hard surface, taking the machine off the duvet. What fixes it categorically: doing sustained heavy work on a fanned machine (Air to Pro is exactly this trade).

2. Memory pressure on 8GB configurations

common

How to confirm it

Activity Monitor → Memory: pressure graph yellow or red, swap in gigabytes, during your normal working set. The 8GB Airs swap early, and heavy swap both slows work and adds heat.

The fix

Fewer simultaneous heavyweights (Chrome tab count is the usual lever). At purchase time, this is why the memory upsell matters more than the chip upsell on Airs.

3. Storage nearly full

common

How to confirm it

macOS slows notably when the SSD has little free space, swap and caches need room. Check About This Mac → Storage: under ~10-15% free is the zone.

The fix

Clear to comfortable headroom; offload media. This one masquerades as chip slowness and costs nothing to fix.

4. A macOS update settling in

occasional

How to confirm it

The days after a major update: Spotlight reindexing and photo analysis run hot, literally, on a fanless machine, and everything feels heavier meanwhile.

The fix

Plugged-in idle hours. Judge the update only after the indexing has finished.

5. Clamshell and hot-room use compound the design

occasional

How to confirm it

Closed-lid on a desk removes the keyboard deck's convection; a 32° room removes the thermal headroom. Both bring the step-down forward.

The fix

Open lid for heavy work; in hot climates, heavy jobs in the cool hours. On a fanless machine the environment IS the cooling system.

Measure it instead of guessing

This machine is the clearest argument for measuring sustained rather than peak: its benchmark scores are excellent and its tenth minute is the truth. Run the deep test, three minutes of held load is exactly where an Air shows its settle point. The number you get is YOUR unit's honest sustained level, in your room, at your desk.

Run the free test

What software cannot tell you here

We cannot read the Air's temperatures or confirm which step of its thermal ladder you are on: Apple exposes none of that to any page.
We cannot see memory pressure or swap; Activity Monitor owns those.

Asked, in people’s own words

Is the MacBook Air M2 throttling a defect?: No, it is the cost of a silent, fanless, thin machine, paid deliberately. Every fanless computer makes this trade. It only becomes a problem when the buyer's workload is sustained-heavy, which is what the Pro line with fans exists for.
Do thermal pads fix M2 Air throttling?: The aftermarket pad mod measurably improves sustained performance by coupling the board to the case better, and it voids expectations of warranty, raises case temperature on your actual lap, and turns the chassis into more of the heatsink than Apple designed. It works; whether it is wise is a different question. We note it exists rather than recommend it.
M2 Air vs M2 Pro for video editing?: For short clips, the Air is genuinely fine, the work bursts. For long exports, the Pro's fan means it holds its level while the Air settles down a step. The deciding question is your timeline length, and a sustained measurement of an Air answers exactly how much the difference would be for you.
Does the M1 or M3 Air have the same issue?: The whole fanless Air line shares the physics; each generation moves the numbers, not the shape. Chip-specific figures deserve chip-specific pages; the measured method for finding YOUR unit's settle point is identical on all of them.

Name the cause. Three and a half minutes, in the tab you have open.

Run the test