Virtual (variable inertia up to 125 kg equivalent)
Noise claim
Not verified from a primary source
Weight
21.5 kg / 47.4 lb
Max rider weight
Not verified from a primary source
Connectivity
ANT+, Bluetooth
These numbers are the manufacturer's claims, recorded as stated, not measurements by us. Database record: Garmin (Tacx) Tacx NEO 2T. Source: garmin.com, checked September 2026.
What independent reviewers found
Findings are paraphrased from the reviewer's own page and credited to them. We ran no tests. Where the reviewer tested an earlier or related model, the reviewed model is named.
DC Rainmaker (Ray Maker)
November 2019
Reviewed model
Tacx NEO 2T
What was measured
Power vs several pedal, crank and hub power meters in Zwift and TrainerRoad, 30x30 ERG repeats
What was reported
Reports the trainer and reference meters staying within a few watts of each other in most data sets, apart from an unsteady first interval; responsiveness was very quick (he says almost too aggressive) and he ranked precision as solid but just behind the Saris Hammer 3 on his 30x30 test.
Found power within about 2% of the pedals and noise around 61 dB; praises the virtual flywheel and quietness but notes no carry handle and a separately bought cassette at launch.
Power on a 5-minute climb and in sprints against Garmin Vector 2S pedals; noise; Zwift climb time
What was reported
Steady climbing differed from the pedals by about 1.2% (4 W) and sprints by roughly 2-4%; the reviewer found the virtual flywheel made climbs and sprints feel harder than on physical-flywheel trainers, and saw 15 W swings in ERG at low wattage. Quotes Tacx's own 57.8 dB noise figure (not measured by the reviewer). Concludes it suits riders prepared for a more challenging ride.
Buy the NEO 2T only at a discount, and only if you want a quiet trainer with strong ERG and are not mainly a Zwift racer. Garmin lists it at $1,099.99 and showed a $899.99 promotional price on 29 Sept 2026. At list prices the KICKR V6 ($1,049.99) costs $50 less, and at its 2026 sale prices much less. Skip it if you race in Zwift: Zwift Insider found its virtual flywheel made climbs and sprints harder than on physical-flywheel trainers.
What reviewers measured
BikeRadar (2019): power within about 2% of Garmin Vector pedals, noise about 61 dB, praise for surface simulation and quietness. No carry handle, and a cassette had to be bought separately at launch.
Zwift Insider (2022): steady climbing about 1.2% (4 W) from Vector 2S pedals, sprints about 2-4% off. Its reviewer found the trainer harder on climbs and sprints, saw 15 W swings in ERG at low wattage, and called it right for riders ready for a more challenging ride. (The 57.8 dB noise figure in that review is Tacx’s own quoted number, not a measurement.)
DC Rainmaker (2019): trainer and reference meters within a few watts in most data sets; very quick, almost aggressive, response; precision just behind the Saris Hammer 3 on his 30x30 test.
All three reviews are from 2019-2022. The specifications on Garmin’s page today are: under 1% claimed accuracy, 2,200 W, 25% grade, virtual flywheel, 21.5 kg.
The trade-off nobody else spells out
The NEO’s inertia is simulated by the motor, not a spinning mass. That is why it is quiet and why ERG is crisp, and also why Zwift Insider felt a penalty in racing efforts. If you want a physical flywheel instead, compare the trainers in the database by flywheel mass.
Who should skip it
Buyers who want Wi-Fi or a bridge for many sensors; and anyone paying list price when the NEO 3M or a V6 is close in cost.
Compared with its closest rivals
Figures are manufacturer claims and store prices from the database; the last column counts independent reviewers whose findings are recorded there.
Read for this page: Tacx NEO 2T manufacturer spec page; DC Rainmaker on Tacx NEO 2T; BikeRadar on Tacx NEO 2T, and 1 more listed above. Checked 29 Sept 2026. SmartBikeWiki runs no tests of its own; every measurement above is the named reviewer’s, paraphrased. Prices change: the store price is the one shown on the manufacturer’s site on the date in the database, which is re-checked every 90 days.