Search for "best palm cooling device" and you get lists written by people who have held one, at most. The devices are genuinely different from each other, and the differences are not the ones the marketing talks about. So instead of a ranking, here are the five numbers that decide whether a palm cooling device does anything for you between efforts, and how the products on the market today compare on each of them.
We make one of these devices. Read this knowing that. Everything below about other products is taken from their own published specifications, and we say where we are guessing.
1. Surface temperature, and whether it stays there
The palm is one of the body's dedicated heat-release surfaces. Under the glabrous skin sit arteriovenous anastomoses, vessels that open under heat load and route warm blood to the surface. A cool surface held against the palm pulls heat out of that blood.
Cool, not cold. Below roughly 14 °C those vessels start to close, and the colder the surface, the less blood arrives to be cooled. (We wrote the long version here: Is colder better for palm cooling?.) So the first number is the surface temperature, and the second question is whether the device holds it. A fixed setpoint near 13 to 15 °C is the design that cannot be set wrong. An adjustable device that goes to 5 or 8 °C can be run into the zone where it works against you, and people who believe colder is stronger will run it there.
Published setpoints: KYLA Performance 14 °C fixed (phase-change plateau). Apex Narwhals about 13 °C. NICE ROCC 13.9 °C fixed. Therabody CryoTherm Palm 8, 12 or 16 °C selectable. CoreTx GO adjustable down to 5 °C. AVA AF Charge Bar 7 to 16 °C drifting, no plateau. Ice and gel packs: 0 °C and rising.
2. Does it fade within a grip?
Every passive device is strong for the first fifteen seconds. The question is the sixtieth. A cold object with nothing behind the surface warms where your hand touches it, and a film of warm material builds between you and the cold. Water-filled bars and gel packs do this within a single grip. A device with a phase-change core and internal conduction paths keeps the surface on its plateau because the whole core supplies the surface, not only the layer under your hand.
This is the number nobody publishes, including us, because it needs an instrumented test rig: surface temperature under a real hand over ninety seconds, repeated for a session. We have modelled it and we are measuring it; until then we describe the mechanism and let you judge the construction. (See The engineering and our methodology.)
3. How much cooling it carries
A palm cooling device is a battery, thermal or electric. Ask how long it delivers, at what rate, before it is spent.
Passive devices carry a fixed amount of cold: for a phase-change device, the mass of material times its latent heat. KYLA Performance carries about 35 to 40 minutes of effective extraction per pair, which covers a two-hour session because you only hold it during rests. A water-filled bar carries a fraction of that and warms out of range in about ten minutes of real use, then needs a refill. An ice-based system with a pump carries hours of cold in a reservoir but has to be tethered to it.
Active (Peltier) devices carry a battery. Therabody publishes 60 minutes cordless. NICE ROCC publishes 90 to 120 minutes. The catch is in section 4.
4. Watts per hand, or why a battery is not the answer
Electric sounds stronger. In a palm cooler it usually is not. A Peltier element in a handheld device has a small battery and a small heat sink, and it can only move a modest amount of heat from your hand before the hot side saturates: a low, steady trickle, held for as long as the battery lasts. A well-built passive device with good contact is limited instead by your own blood flow, which for a heat-stressed athlete is a much larger number. It cannot hold that for hours, but a training session does not ask it to. Passive or battery-powered: what electricity buys you goes through this properly.
The only devices that combine high sustained watts with long duration are tethered: an ice-and-pump mitt (single hand, a console on the floor) or a mains-powered bench chiller. Excellent in a lab or a locker room. Not in a gym bag.
5. Weight, hands, and whether you will actually use it
The last number is the one that decides everything: will it be in your bag and in your hand between sets, every session? Weight per cooled hand, whether it does one hand or two, whether it needs a freezer, a socket, ice, or nothing.
Two hands is not double one hand, because both palms draw on a shared central blood supply; a pair delivers roughly 1.6 times a single. But a single-hand device still leaves half the surface unused. (Why two hands is 1.6, not 2.)
The current products, on published numbers
| Product | Type | Hands | Weight | Surface temperature | Power / recharge | Published run time | Price (published) |
|---|---|---|---|---|---|---|---|
| KYLA Performance | passive, phase-change core, patent-pending internal system | 2 | ~0.44 kg per handle | 14 °C fixed | fridge, 5 h | ~35–40 min of contact, covers a 2 h session | 3,295 kr (pair) |
| Apex Narwhals | passive, copper heat pipes to a separate pack | 2 | 1.4 kg per unit | ~13 °C | freezer/fridge | ~2 h at low output | $199 each |
| Therabody CryoTherm Palm | active Peltier, handheld | 2 | 517 g per unit | 8 / 12 / 16 °C | battery | 60 min cordless | $400 |
| NICE ROCC | active Peltier, dome | 2 | ~2.0–2.3 kg system | 13.9 °C fixed | battery, 64 Wh | 90–120 min | $975 |
| CoolMitt | active, ice water + pump, mitt | 1 | console, several kg | ~12 °C water | ice; 6 h per fill, 8–10 h recharge | hours (tethered) | ~$995 |
| AVA AF Charge Bar | passive, water-filled aluminium bar | 2 | ~0.91 kg | 7–16 °C, drifts | tap water/fridge | ~10 min then refill | $74.95 |
| Kuhler | active Peltier, single | 1 | 313 g | ~15 °C | battery | not published | $199–289 |
| CoreTx GO | active, chilled water jet, bench | 2 | ~7 kg | adjustable to 5 °C | mains | indefinite | ~$3,000 |
| Gel pack / Tempsicle-type | passive gel in fabric | 1 | ~0.5 kg | drifts from freezer temp | freezer | tens of minutes | $69–89 |
Prices and figures checked on 3 September 2026, on each maker's own page. Tell us if one has changed.
Prices and specifications from each maker's site as of July 2026; check before you buy. Weights for CoolMitt and CoreTx are estimates.
How we would read that table
If you train in a gym and want something that works between sets without a socket, a freezer or a bucket, you are choosing among the passive devices and the handheld Peltiers. Among those, the questions are: does it hold near 14 °C (fixed plateau beats drift and beats adjustable-to-cold), does it fade within a grip (internal conduction beats still water beats gel in fabric), and does it carry enough for a session. If you want hours of cooling in one place and do not mind the tether, the ice mitt and the bench chiller are a different category and a different price.
We built KYLA Performance for palm cooling between sets, and to win the first set of questions: a fixed 14 °C plateau, a patent-pending internal system that puts the whole core behind the surface so it holds between sets, two hands, no battery, no freezer. In our heat-transfer model and in our own testing, it removes more heat per rest than anything else we have measured; how we measure, and what we have not measured yet, is on the methodology page. What it does not do is run for hours in the sun; that is what the tether buys you. KYLA Performance →