Amazfit Cheetah 2 Ultra Review: Strong HR, Steep Premium

What it is

The Amazfit Cheetah 2 Ultra is a performance running watch positioned at the upper end of Amazfit's 2026 lineup, sitting £150 above the Cheetah 2 Pro that launched just four weeks earlier. It targets serious runners and hybrid athletes who want accurate biometric tracking without paying Garmin Fenix or Apple Watch Ultra prices. Based on available pricing relative to the Pro, it sits in the mid-to-upper price tier, likely landing around £400-plus depending on market. The key question from the moment it launched was simple: does that £150 premium buy anything real?

Key specs

• GPS/GNSS: Multi-band GNSS (chipset details not confirmed in available sources)
• Battery life: Not confirmed in available sources
• Sensors: Wrist optical PPG for HR and HRV, SpO2 (blood oxygen via optical sensor), SmO2 (muscle oxygen saturation sensor), barometric altimeter, skin temperature
• Display: Not confirmed in available sources
• Weight: Not confirmed in available sources
• Water resistance: Not confirmed in available sources

The standout hardware addition over the Pro appears to be the SmO2 sensor, which measures muscle oxygen saturation via optical means at the wrist. That is the headline differentiator Amazfit is charging a £150 premium for.

Performance in the real world

This is where the Cheetah 2 Ultra actually impresses. Three separate HR accuracy tests conducted by The5kRunner produced clean results across difficult conditions.

In a HYROX simulation at 33 degrees Celsius, a genuinely brutal environment for wrist optical PPG sensors, the Cheetah 2 Ultra showed near-zero bias against three reference devices: the Whoop MG, the Garmin HRM-600 chest strap, and the Polar Verity Sense optical arm band. Getting a clean result in that kind of heat and intensity is not trivial. Wrist optical sensors typically struggle when sweat, movement, and high ambient temperature combine, so this result matters.

In a road bike sprint HR test, the Cheetah 2 Ultra returned limits of agreement of plus or minus 3 bpm against reference devices. For context, the Fitbit Air tested simultaneously in the same protocol came back at plus or minus 6 bpm. That is a meaningful real-world difference. When you are sprinting at threshold or above, a 6 bpm error can push you into the wrong training zone. A 3 bpm error is close enough to trust for most training decisions.

Three tests, three clean results. That consistency is the kind of track record that matters more than any single data point. It suggests the optical PPG implementation here is genuinely good, not just lucky on one test day.

On the SmO2 front, the sensor exists and is a differentiator, but available sources do not provide enough real-world validation data to say confidently how accurate or useful it is in practice. SmO2 at the wrist is technically challenging and remains a relatively unproven metric compared to chest-based or finger-based devices from companies like Moxy. Buyers should treat it as a bonus feature to explore rather than a primary reason to choose this watch.

Sleep tracking quality, GPS accuracy in the field, and app ecosystem depth are not addressed in the available sources with specific numbers. The Zepp app that powers Amazfit devices has historically been adequate for training load and recovery tracking but lacks the depth of Garmin Connect or the coaching sophistication of platforms like Polar Flow. That gap likely persists here.

Who it's for / who should skip it

The Cheetah 2 Ultra is for the serious runner or hybrid athlete who trains in high-intensity, high-sweat environments and needs reliable wrist HR data they can actually trust. The HYROX test result alone makes this worth considering for CrossFit athletes, obstacle racers, and anyone doing functional fitness alongside their running. If HR accuracy in tough conditions is your primary concern and you want to stay under Garmin Fenix 8 pricing, this is a legitimate option.

Skip it if you are already on the Cheetah 2 Pro. The spec sheets are almost identical, the differences took a reviewer ten minutes to identify, and £150 is a lot of money for an SmO2 sensor that lacks independent validation data. Skip it also if you are deep in the Garmin or Apple ecosystem and rely on third-party app integrations, Garmin Connect IQ, or watchOS compatibility. Amazfit's Zepp ecosystem does not match those platforms for breadth.

If you are choosing between the Cheetah 2 Ultra and something like the Garmin Forerunner 965, the Garmin wins on ecosystem maturity, GPS track record, and training analytics depth. The Amazfit wins on HR accuracy in intense conditions based on current test data, and likely on price. That trade-off is real and worth thinking through honestly.

Verdict

The Amazfit Cheetah 2 Ultra delivers genuinely impressive optical HR accuracy across multiple tough real-world tests, and that is not a small thing. The £150 premium over the Cheetah 2 Pro is hard to justify unless you specifically want the SmO2 sensor and understand its current limitations. Buy it over the Pro only if you are coming from outside the Amazfit lineup entirely.