17 reviewsAll reviews →
Buying guide · Keep it running

Portable power station for CPAP machine backup: nights per charge, switchover and each maker's medical stance

As an Amazon Associate I earn from qualifying purchases.

At a glance6 picks
  1. Best for most CPAP usersBLUETTI Elite 300 Under $1,200Buy Check price on Amazon
  2. Lightest CPAP backup for a night or twoAnker SOLIX C1000 Gen 2 Under $600Buy Check price on Amazon
  3. Best for a CPAP on the tightest budgetEcoFlow DELTA 3 Classic Under $500Buy Check price on Amazon
  4. Most of a week for lessBLUETTI Elite 200 V2 Under $900Buy Check price on Amazon
  5. Best for week-long outages with a generatorAnker SOLIX F3000 Under $1,700Buy Check price on Amazon
  6. Best if the fridge comes firstJackery HomePower 3600 Plus Under $1,900Buy Check price on Amazon
The verdict

The BLUETTI Elite 300: about seven nights of CPAP per charge by our arithmetic, a switchover timed at 8 ms, and no power-saving default that cuts a small overnight load. For CPAP users facing outages of a day to a week.

Most CPAP owners shop by battery size. That turns out to be the easy part, because a CPAP sips power. What goes wrong at 3 a.m. is a station that decides nothing is plugged in and switches its outlets off. Or one that spends a big share of every night keeping itself awake. Porchlight's pick for a CPAP through an outage is the BLUETTI Elite 300. By our arithmetic it gives about seven nights per charge. StorageReview timed its switchover at 8 ms, quicker than BLUETTI's own 10 ms claim. And BLUETTI leaves its ECO cut-off disabled out of the box, so a light overnight load keeps its power. The catch is size and sound. BLUETTI's own manual suggests two people to lift a box this heavy. Nobody has measured its fans at a CPAP's tiny load, though flat out they are louder than rivals. We'd keep it across the bedroom, not on the nightstand.

Check price on AmazonOpens Amazon.com · Prices change often

What changes the answer for a CPAP or oxygen user

Every figure here assumes the heated humidifier is off. CPAP retailers say heat raises the draw a lot, so with it on, expect fewer nights. An oxygen concentrator is a different job, and we answer it below. The Elite 300 has energy to spare for a fridge as well as the CPAP. But if the fridge is your first worry and the CPAP rides along, buy for the fridge: that's the Jackery further down. If your outages run past a week, refilling matters more than battery size, and the Anker F3000 with a generator is the answer. If you rent an apartment and the station must cover the fridge and Wi-Fi as well as the CPAP, our apartment backup guide picks the Anker SOLIX C2000 Gen 2. For the rest of the house, the home backup power calculator works from testers' measurements to show how long your essentials would run.

Best for most CPAP users: BLUETTI Elite 300

BLUETTI Elite 300

Price band: under $1,200

BuyPower stationRead the review

BLUETTI promises a 10 ms switchover, and that undersells it. StorageReview put an analyzer on the outlet and caught it going to battery in 8 ms. No other station its size in our evidence has been timed. I'll take a measured 8 over a printed 10 every time.

Two US units delivered 2,670 to 2,810 Wh, about nine tenths of the label. Its idle draw is 15 W, a figure three separate measurements landed on. With a CPAP running, that idle is a third of each night's energy. That's why it lands at about seven nights rather than ten. A CPAP leaves most of that battery free, so there is room for the fridge in a shorter outage, though nobody has yet run a fridge on this unit.

BLUETTI's manual settles two night-time worries. ECO, its power-saving cut-off, is disabled from the factory, so the outlet won't decide your CPAP isn't there. And if the battery runs flat, the outlets come back on their own once the grid returns and the battery recovers a little.

Close-upElite 300, close-up
Elite 300Amazon listing

The one real catch is the fans. One tester measured 53 dB at full load and 49 dB while fast-charging, louder than rivals. Nobody has measured it at a CPAP's load. If the grid returns at 2 a.m. and it starts refilling, you may hear it. The manual's Silent charging mode trades speed for quiet. With no wheels and a two-person weight by BLUETTI's own reckoning, set it up once and leave it.

Buy it for a week of nights on a switchover a lab actually timed.

Check price on AmazonOpens Amazon.com · Prices change often

Lightest CPAP backup for a night or two: Anker SOLIX C1000 Gen 2

Anker SOLIX C1000 Gen 2

Price band: under $600

BuyPower stationRead the review

Anker advertises "20dB Operation" without saying at what load. That's like selling a quiet car by measuring it with the engine off. The numbers that matter hold up better. Two labs got 875 Wh or a shade under out of it. Anker's feature list claims 9 W on standby; a solar installer measured about 14 W, twice. That's still low for this field. It works out to about two and a half nights of CPAP, from a 24.9 lb box one person can carry.

Anker's switchover claim of under 10 ms is untimed. One owner pulled the wall plug on six running machines, a 3D printer mid-print among them, and nothing stopped. The solar installer also saw its outlets return unprompted when an emptied battery began taking charge again.

The catch is that two and a half nights is the ceiling. There's no expansion port, so a three-day ice storm means refilling it from somewhere. We had only Anker's troubleshooting guide for this model, not its full manual, so we can't tell you its power-saving settings.

Buy it for a night or two, carried by one person.

Check price on AmazonOpens Amazon.com · Prices change often

Best for a CPAP on the tightest budget: EcoFlow DELTA 3 Classic

EcoFlow DELTA 3 Classic

Price band: under $500

BuyPower stationRead the review

This is the least you can spend to keep a CPAP running for a night or two. It is also the box we pick for homes whose power blinks off often and briefly, wanting the router and desk to ride through and the battery to refill fast; our DELTA 3 Classic review covers that buyer. EcoFlow's switchover claim is under 10 ms, and no one has put an instrument on it. Three testers did cut the power under lights, a TV and an aquarium's pumps, without a flicker. One tester says his lamp shows a flicker at 15 ms or more. Two labs measured about 940 Wh usable and idle of 13 to 17 W. That's about two and a half nights of CPAP, like the Anker, from a lower price band. EcoFlow claims under 30 dB at a moderate load; one tester read 39 to 40 dB under a far heavier one.

DELTA 3 Classic
DELTA 3 ClassicAmazon listing

The catch is a timer. EcoFlow's US manual says the AC outlets switch off when idle, warns a fridge can trigger it, and gives no default length. Follow the manual and set the AC timeout to "never" in the app before your first outage. With no car socket, the CPAP runs through the AC outlets.

Buy it for a night or two of CPAP on the tightest budget, with the timer off.

Check price on AmazonOpens Amazon.com · Prices change often

Most of a week for less: BLUETTI Elite 200 V2

BLUETTI Elite 200 V2

Price band: under $900

BuyPower stationRead the review

Four testers measured the Elite 200 V2's usable energy between 1,910 and 2,030 Wh. That's about as close as independent tests get. BLUETTI says it uses about 10 W running itself. Testers got 9.5 W to 14.7 W depending on method, so BLUETTI earns the credit. That adds up to five to six nights of CPAP.

Its 15 ms switchover claim is untimed. But it kept desktop computers running through grid cuts in three separate tests. One tester found its slowest charging mode quiet enough to vanish, which counts if it refills by the bed.

The catch is one default. BLUETTI's manual ships ECO mode on, cutting output after a stretch of low load. The same manual wants ECO disabled wherever the load is a small device. A CPAP is a small device, so switch ECO off the day it arrives. At 53.35 lb, it has no wheels and no light.

Buy it if your budget stops short of the Elite 300 and five nights will do.

Check price on AmazonOpens Amazon.com · Prices change often

Best for week-long outages with a generator: Anker SOLIX F3000

Anker SOLIX F3000

Price band: under $1,700

BuyPower stationRead the review

Anker claims a 20 ms switchover. Jasonoid's oscilloscope measured 9.8 ms, half the claim, while a stressed PC rode through three cuts. A maker underselling its own box is rare enough to mention.

Refilling is what earns the F3000 its place. Plug a 120 V generator into its TT-30 charging cable and it will draw up to 3,600 W. Testers confirm a refill that way takes under an hour. A short generator shift each day keeps the CPAP going as long as the fuel lasts. Carbon monoxide is why the generator lives outdoors and the battery is the thing in your bedroom.

The catch is idle draw nobody can pin down: 20 W to 35 W across testers. One tester even got different answers a year apart. That swings a full charge between five and six and a half nights. At 91.5 lb it rolls on wheels, the only way it reaches a bedroom. We couldn't get its US manual, so we can't say which settings affect a small overnight load.

Buy it for outages measured in weeks, paired with a generator.

Check price on AmazonOpens Amazon.com · Prices change often

Best if the fridge comes first: Jackery HomePower 3600 Plus

Jackery HomePower 3600 Plus

Price band: under $1,900

BuyPower stationRead the review

This is our pick for keeping a home's essentials running, and it is the station for a home whose first worry is food, lights and Wi-Fi, with the CPAP riding along. For a CPAP alone it is more box than you need. Three capacity tests agreed at 3,201.32 to 3,270 Wh, the most here. In a lab it kept a full-size fridge going from one morning to the evening of the next day. A CPAP alone gets six to eight nights; with the fridge too, a bit less than that fridge run. Its idle draw came in at 21 W in one test and 34 W in another. On the worse reading, the station's own upkeep outstrips the CPAP's.

The catch is a default that seems made to catch CPAP owners. Jackery's manual turns the outputs off after 12 hours below 25 W, and it ships switched on. A CPAP plugged in through the day shows twelve hours of next to nothing. By bedtime, the outlets may be off. The manual's way out is a three-second press on the AC and main power buttons at once. Jackery's 10 ms switchover claim is untimed; a running computer stayed up through one cut.

Buy it if the fridge matters more than the CPAP. More in our HomePower 3600 Plus review.

Check price on AmazonOpens Amazon.com · Prices change often

New this year, and what we're watching

The Elite 300 went on sale in March 2026, making our pick the newest station here. A box that new usually means waiting, but not this time. In eight months it has gathered more of what a CPAP user needs than models years older. It has a timed switchover, three matching idle tests, two capacity tests on US units and a six-month owner. That owner reports the app often loses the unit in standby. It also didn't wake for solar after powering down overnight. That's one person's report, not a pattern, and we're watching for more.

BLUETTI has also just put an Apex 200 on its US store. Nobody in our evidence has measured the Apex 200 itself, so there's nothing to weigh and no reason to hold off for it.

How the picks were decided

Ten models went into this page, nine of them read against the maker's own manual, drawn from 589 sources this site has kept in full. A station made the list only with measured usable energy and idle draw on the exact US model. At least one of those measurements had to come from a tester who isn't the maker.

How many nights of CPAP will a power station give me?

Divide the energy labs measured out of the battery by one night's cost. That cost is the CPAP plus the station's own idle draw, over eight hours. One tester metered a ResMed AirSense 10 with the humidifier off at around two dozen watts on average. He suggested planning on thirty, and our arithmetic uses that.

We add idle draw because at a load this small, the inverter's appetite is most of the loss. Anker itself says efficiency falls at light loads. Labs measured usable energy at much heavier loads, so treat our nights as a best case and round down. Each pick's nights are in the table below.

Does switchover time matter for a CPAP?

Only if the CPAP lives on the station nightly, with the station plugged into the wall. We found no source on how a ResMed reacts when power blinks. So we can't name a figure your machine is sure to tolerate. We can say which figures are real. Only four stations in this field have a measured switchover, and two are here: the Elite 300 at 8 ms and the F3000 at 9.8 ms. The other two are the BLUETTI Apex 300 and a big Jackery measured through its transfer switch. Every other number on this page is the maker's own.

Will the station switch itself off under a CPAP?

Ask before the outage, not during it. Power-saving modes cut outlets that see little load, and a CPAP idle all day looks like nothing at all. The Elite 300's ECO mode ships off. The Elite 200 V2's ships on, and its manual says to switch it off. EcoFlow's manual says to set the DELTA 3 Classic's AC timeout to "never" in the app. Jackery's manual shows how to turn off the 3600 Plus's Energy Saving Mode with the buttons. We lacked full manuals for the C1000 Gen 2 and F3000, so we can't name their defaults.

What do the makers say about medical use?

Nobody promises anything, so read the wording exactly. EcoFlow's DELTA 3 Classic FAQ says it is "not recommended for critical medical devices such as ventilators." Jackery says, brand-wide, that it doesn't recommend its stations for medical or life-support equipment. Its 3600 Plus manual adds that it does not support 0 ms switching. BLUETTI's manuals warn against servers and workstations that need a high-performance UPS. We found nothing from BLUETTI on medical devices either way. From Anker we found no statement at all. Silence is not permission. None of these is a medical device, and none of the four makers will stand behind its box as one.

Can a power station run an oxygen concentrator overnight?

Perhaps for a single night on the largest box here, and we can't promise even that. The research behind this page assumes a stationary concentrator draws a few hundred watts, many times a CPAP. Nobody in our evidence measured one, so read the label on yours. The Jackery holds the most measured energy here, which makes it the only station we'd even try. Jackery is also the maker most firmly against medical use. If a lapse in oxygen would put you in danger, no station here is a plan on its own. Ask whoever supplies your equipment what they provide for outages.

The picks side by side

StationBest forNights of CPAP per chargeSwitchoverWeightBuy
BLUETTI Elite 300Most CPAP usersAbout 78 ms, measured57.98 lbView on Amazon
Anker SOLIX C1000 Gen 2A night or two, carried by one personAbout 2½Under 10 ms, claimed24.9 lbView on Amazon
EcoFlow DELTA 3 ClassicA CPAP on the tightest budgetAbout 2½Under 10 ms, claimed26.7 lbView on Amazon
BLUETTI Elite 200 V2Most of a week for less5 to 615 ms, claimed53.35 lbView on Amazon
Anker SOLIX F3000Week-long outages with a generator5 to 6½9.8 ms, measured91.5 lbView on Amazon
Jackery HomePower 3600 PlusThe fridge first, CPAP along6 to 8, CPAP alone10 ms, claimed77.16 lbView on Amazon

Our pick for CPAP backup: BLUETTI Elite 300

For a CPAP through an outage, we'd buy the BLUETTI Elite 300. It holds about a week of nights. Its switchover is measured rather than promised. And nothing in its defaults switches your CPAP off while you sleep. If the fridge worries you more than the CPAP, our home backup guide picks the Jackery HomePower 3600 Plus instead.

A CPAP asks for very little, and the Elite 300 keeps giving it that, night after night. Buy it, set it up once, and let it get on with the night.

Check price on AmazonOpens Amazon.com · Prices change often

The evidence in full

Research count: 10 models weighed; 9 maker manuals among them; across this site's research, 589 sources kept in full.

Every station considered, with the nights working

Each night is an eight-hour night at thirty watts for the CPAP (a metered ResMed AirSense 10 averaged around two dozen with the humidifier off; thirty was the tester's own planning figure), plus the station's measured idle draw with the AC on. Nights are measured usable energy divided by that, and they are our arithmetic, not anyone's test. Every usable-energy figure came from a load far heavier than a CPAP, where inverters are more efficient, so these are upper estimates.

StationUsable energy, measuredIdle draw with AC on, measuredNights per chargeSwitchoverCall here
BLUETTI Elite 3002,670 Wh (StorageReview); 2,810 Wh (HOBOTECH)15 W (HOBOTECH, The Solar Lab, Bren Unplugged)7.4 to 7.88 ms to battery, measured (StorageReview); 10 ms claimedBuy, overall pick
Jackery HomePower 3600 Plus3,201.32 Wh (CNET); 3,240 Wh (HOBOTECH); 3,270 Wh (Adventure Gear TV, Jackery-sponsored)21 W (HOBOTECH, display-based); 34 W (The Solar Lab)6.3 to 7.810 ms claimed; computer ride-through untimed (HOBOTECH)Buy
BLUETTI Elite 200 V21,987.98 Wh (CNET); 1,910 Wh (StorageReview); 1,990 Wh (HOBOTECH); 2,030 Wh (ReeWray)9.5 W (The Solar Lab); 14.7 W (HOBOTECH, includes charge losses); 10 W claimed5.6 to 6.315 ms claimed; three PC ride-throughs untimedBuy
Anker SOLIX F30002,645.91 Wh (CNET)20 W (Jasonoid); 20.5 W (The Solar Lab); 35 W (HOBOTECH)5.1 to 6.59.8 ms measured (Jasonoid); 20 ms claimedBuy
Anker SOLIX C1000 Gen 2870.08 Wh (CNET); 875 Wh (Jungle River)13.6 W and 14 W (Minute Man Solar, two videos)About 2.5Under 10 ms claimed; one owner's ride-through untimedBuy
EcoFlow DELTA 3 Classic947.47 Wh (CNET); 940 Wh (The Solar Pit)13 W (The Solar Pit, screen-based); 17 W (The Solar Lab)2.5 to 2.8Under 10 ms claimed; ride-throughs untimedBuy
Anker SOLIX C2000 Gen 21,812.04 Wh (CNET)9 W (Adventure Gear TV, Anker-supplied unit); 18 W (The Solar Lab)4.7 to 5.810 ms claimedConsidered
BLUETTI Apex 3002,390 Wh (StorageReview); 2,480 Wh (There's A Trick For That)16 W (There's A Trick For That); 20 W (The Solar Lab)6.0 to 6.716 ms measured in its standard mode (StorageReview); 0 ms claimed only in one mode on two outletsConsidered
EcoFlow DELTA 3 MaxNot measured27 W (The Solar Lab)Not worked outUnder 10 ms claimedSkip
Jackery Explorer 1000 v2Not measuredNot measured; 18 W claimedNot worked out20 ms claimedSkip

The Solar Lab measured 13.8 W idle on the C1000X Gen 2, the big-box sibling of the C1000 Gen 2 with four AC outlets instead of five. It agrees with Minute Man Solar but is not this model's result.

A route nobody has tested: running a CPAP from a car socket through the CPAP maker's own DC converter skips the inverter. HOBOTECH measured the Elite 300 at 3.75 W idle with only DC on, against 15 W with the AC on, which would stretch its nights considerably. One CPAP retailer recommends a DC converter over an AC inverter. We found no test of a CPAP on any station's DC output, and only your CPAP's maker can say which converter is safe for it.

No fridge run exists for the Elite 300. Its room for a fridge alongside the CPAP is our reading of its measured usable energy, not a test; the Jackery's fridge run is CNET's lab result.

The DELTA 3 Classic earns its place here for two buyers our other pages also serve. For a CPAP, it is the cheapest way through a night or two. For homes with short, frequent outages that want the router and desk to ride through and the battery to refill fast, it is our pick too, as its review explains.

The four measured switchovers in the field

  • BLUETTI Elite 300: 8 ms to battery, and about the same on the way back to the grid, on a Quarch analyzer with a workstation load (StorageReview, US unit, Aug 2026).
  • Anker SOLIX F3000: 9.8 ms on an oscilloscope; a PC under stress survived three cuts (Jasonoid, US unit, Jul 2025).
  • BLUETTI Apex 300: 16 ms in its standard mode, captured at two loads. Its 0 ms mode was not captured (StorageReview, 2026-11-10).
  • Jackery Explorer 5000 Plus, through Jackery's Smart Transfer Switch: a whole-circuit switchover, not an outlet test (The Solar Lab and others). That station is not among the models this page weighed.

No tester timed any station in a 0 ms mode.

Each maker's stance, in its own words

MakerWhere it says itWhat it says
EcoFlowDELTA 3 Classic US FAQ and manual"Not recommended for critical medical devices such as ventilators." The manual adds that the grid must supply more than the loads draw, or the station cannot act as a UPS.
JackeryFAQ on the Explorer 5000 Plus page (brand-wide); HomePower 3600 Plus manual; Explorer 1000 v2 FAQDoes not recommend its power stations for medical or life-support equipment. The 3600 Plus "does not support 0 ms switching" and should not feed equipment that needs an uninterrupted supply, such as data servers or workstations. The Explorer 1000 v2 is "not a 0 ms (no interruption) switching UPS."
BLUETTIElite 200 V2, Elite 300 and Apex 300 manualsNot for servers or workstations that need a high-performance UPS. No medical statement found.
AnkerC1000 Gen 2 page and guide, C2000 Gen 2 manual, F3000 pageNo medical or life-support statement found. The C1000 Gen 2 guide says every AC outlet supports UPS.

Settings that can cut a small load, by manual

  • BLUETTI Elite 300: ECO off by default. If the battery runs flat, AC output returns once the grid is back and the battery recovers.
  • BLUETTI Elite 200 V2: ECO on by default, and the manual wants it disabled for small devices. System Switch Recovery restores output after a flat battery.
  • BLUETTI Apex 300: ECO on by default; the manual advises disabling it for small loads.
  • EcoFlow DELTA 3 Classic: AC timeout on, no default length given; the manual says set it to "never" in the app. Powers on again when charging resumes, and Port Memory restores the outputs.
  • EcoFlow DELTA 3 Max: AC idle timeout; the manual says set it to "never" for fridges and medicines.
  • Jackery HomePower 3600 Plus: outputs off after 12 hours below 25 W, on by default; disable it with the AC and main power buttons.
  • Anker SOLIX C2000 Gen 2: AC turns off after 15 min with no plug detected; on by default, and the manual says it can be disabled.
  • Anker SOLIX C1000 Gen 2 and F3000: their full US manuals were not available to us. One owner watched the C1000 Gen 2's outlets come alive again without help as its drained battery recharged, and MacRumors notes its AC has to be switched on by hand.

Testers and their methods

TesterHow it measuresWho supplied the unit
StorageReviewKill-A-Watt for energy in and out; Quarch analyzer for switchoverNot stated
JasonoidOscilloscope on the outlet during a grid cut; idle read from the screenNot stated
CNET LabsUsable energy from a fixed resistive load logged to empty; round-trip efficiency, which includes charging losses, published separately; unit market not always statedNot stated
HOBOTECHIdle read either from the screen percentage over hours or from the energy needed to recharge, which adds charging lossesMaker discount or affiliate codes on several videos
The Solar LabSays it tests idle draw in every review; method not always statedBought its Elite 200 V2 anonymously
ReeWray Outdoors, Ruby's Adventures UnpluggedFull discharges through the inverter; generator runsMaker discount or affiliate codes
Adventure Gear TVHeat-gun discharge; metered idle over a long runJackery sponsored its 3600 Plus video; Anker supplied its C2000 Gen 2

Every result on this page comes from a US unit. Measurements made on UK, European and Australian versions were left out of every figure here.

Disagreements that don't change the call

Idle draw shows a spread on every station, and most of it comes from method: screen percentages are coarse, and recharge-energy readings include charging losses. The F3000 is the exception worth noting, because The Solar Lab's own two runs a year after its first test disagreed by more than double, which points to firmware or the unit itself as well as method.

BLUETTI claims the Elite 300 lasts 6,000+ cycles, while HOBOTECH quotes at least 4,000. Neither figure changes nights per charge.

Jackery's manual says the 3600 Plus charges at subfreezing temperatures, while a third party reports a Jackery FAQ setting the floor at freezing. A bedroom is never that cold, so it doesn't matter here.

CNET's spec box for the DELTA 3 Classic lists ports that don't match the Classic, but its product image and charge time do, and two other Classic tests agree, so we treat CNET's figure as the Classic's.

Skip these, and also considered

The Anker SOLIX C2000 Gen 2 gives about five nights, but for a CPAP it lost to the Elite 200 V2. It has one capacity test, idle readings of 9 W and 18 W that disagree by double, and unconfirmed reports that its app sends no outage alert. It is a buy for a different reader: someone in an apartment or small home who carries the battery alone and needs the fridge and Wi-Fi covered through a day. At 41.7 lb it suits that job, and both our apartment backup guide and our home backup guide pick it there; see the C2000 Gen 2 review.

The BLUETTI Apex 300 gives about six nights and has a measured 16 ms. But it weighs 83.78 lb, has no USB or car socket without an extra hub, ships with ECO on, and is built for a house wired through a transfer switch. That buyer is covered in the Apex 300 review.

Skip the EcoFlow DELTA 3 Max. It has one tester and no lab usable-energy figure, so we couldn't work out its nights honestly. Its 27 W idle is the highest of the boxes in its size class here, and it weighs 44.8 lb.

Skip the Jackery Explorer 1000 v2 too. It is light at 23.8 lb and has a car socket. But nobody has measured its usable energy, idle draw or switchover; Jackery claims 18 W standby and a 20 ms switchover, and its maker advises against medical use. There is too little measured to back it.

Sources

Best for most CPAP usersBLUETTI Elite 300 Under $1,200 Check price on Amazon
↑ ↓ to move↵ to open