Buying guide

Choosing an e-bike for hills

Gradient is where cheap e-bikes fall apart. Torque, gearing, cooling and brakes matter more than headline watts — and sixteen of the twenty-two bikes we cover publish the number that settles it.

Published 2026-08-21 · updated 2026-08-22 · 12 min read

The short version

Torque in newton-metres predicts climbing; watts do not. Published figures across the twenty-two bikes run from 40 Nm to 220 Nm, and six bikes publish none at all — all five AMYETs and the VTUVIA SN100.
Watts are electrical input. Newton-metres are how hard the motor turns the wheel — and on a mid-drive, the gears multiply that torque again.
Compare rated power, never peak. Peak output is a burst figure with no bearing on a climb that lasts more than about thirty seconds.
A hub motor held at low road speed and high load gets hot, and a hot motor has its current cut back by the controller to protect itself.
Gear count matters even on a hub bike, and matters enormously on a mid-drive: seven speeds against SRAM NX eleven is a different climbing tool. There are five mid-drives here now, from three brands, not two from one.
Plan the descent as carefully as the climb. On a 110 lb bike, brake capacity and published rotor size matter as much as motor output.

Why watts do not predict climbing

Almost every budget e-bike in this market claims 1000W or more, and almost none of them climb the same way. The watt figure is not why they differ.

Power is the rate of doing work. Torque is turning force. On a gradient at low road speed, what gets you up the hill is turning force at the wheel — and two motors rated at the same wattage can produce very different torque depending on winding, magnet count, the reduction gearing inside the hub, and how the controller is programmed. This is why a mid-drive rated at 250W can out-climb a hub motor rated at four times that.

There is a cleaner way to hold the two ideas apart. Watts describe what goes in: volts multiplied by amps, the electrical input the controller is willing to push at the motor. Newton-metres describe what comes out: how hard the motor twists the thing it is bolted to. A brand that publishes only the first number has told you the size of the tap and nothing about the pipe.

Sixteen of the twenty-two bikes we cover publish a torque figure. Six do not.

40 → 220 Nm

published torque, lowest to highest

The ladder, from Velotric's road commuter to EUNORAU's flagship climber, is worth committing to memory:

Velotric T1 ST Plus — 40 Nm, from a 350W rear hub. The lowest figure on this site, in the lightest bike on this site, and on the narrowest tires we cover.
EUNORAU META26 1.0 — 55 Nm, from a 500W rear hub. The lowest of EUNORAU's five.
VTUVIA Zeal LT7 — 65 Nm, also 500W, also a rear hub.
Velotric Nomad 1 Plus, Packer 1, Discover 2 and Discover 3 — 75 Nm each, from 750W rear hubs, on bikes rated to carry 440 lb.
VTUVIA SF20H and VTUVIA SX20 — 85 Nm each, from the same 750W rear hub.
EUNORAU FLASH Lite 2.0 — 92 Nm, from a 52V 750W rear hub in its base configuration; 184 Nm in the dual-hub AWD option, which is two motors added together.
VTUVIA Gemini — 100 Nm, rated 1000W, position not stated. Above the 92 Nm of EUNORAU's FLASH Lite rear hub, which is the most any bike here states from a confirmed single hub motor, and above the SF20H and SX20 at 85 Nm.
Velotric Discover M — 100 Nm, from a 500W VeloCore mid-drive through a 9-speed 11–46T. The same number as the Gemini, arrived at the other way.
EUNORAU SPECTER-S 3.0, EUNORAU SPECTER-ST 2.0 and VTUVIA FMB — 160 Nm each, all three from a 1000W Bafang G510 mid-drive.
EUNORAU FLASH 2.0 — 220 Nm, from a 1000W Truckrun mid-drive. The highest figure on this site by a wide margin.

The gap is worth sitting with. AMYET's G60 claims a rated figure double the Zeal LT7's, and we still cannot tell you which of the two pulls harder up a 12% ramp, because only one of them tells you. VTUVIA's Gemini is the tidiest counter-example: also rated 1000W, and it publishes 100 Nm. Our motor wattage guide takes the power figures apart in more detail.

Rated versus peak, on a climb that lasts

Two power figures appear on most listings and only one of them describes hill riding.

Rated power is what the motor can hold more or less continuously without exceeding its thermal limits. Peak power is the short burst the controller will allow before something gets too hot — useful for pulling away from a junction or cresting a short rise, and gone within seconds.

A climb is not a burst. If your route has half a mile at 8%, you are asking the motor for sustained output for two or three minutes at a road speed low enough that airflow over the hub is minimal. The rated figure governs that period. The peak figure has nothing to say about it.

VTUVIA publishes 1200W peak against 750W rated on both the SF20H and the SX20, 750W peak against 500W rated on the Zeal LT7, 1500W peak against 1000W rated on the Gemini, and 1300W peak against 1000W rated on the FMB; for the SN100 it publishes no peak at all. AMYET publishes 1500W peak against 1000W rated on the G60, the EB26 and the K10, and a 3000W combined peak against 2 × 1000W rated on both the S8 and the Ares. EUNORAU publishes no peak figure for the META26, the FLASH, the FLASH Lite or the SPECTER-ST at all; for the SPECTER-S it publishes something more useful instead — a 48V 30A controller, which is about 1,440W of electrical input at full draw. Velotric publishes a 600W peak against a 350W rating on the T1 ST Plus, 1100W against 750W on the Discover 3 and 960W against 500W on the mid-drive Discover M, and no peak figure at all for the Nomad 1 Plus, the Packer 1 or the Discover 2.

When comparing bikes for hilly ground, line up rated against rated and ignore the headline.

Heat, thermal cutback and what bogging down feels like

A hub motor is an electric motor sealed inside a wheel. It sheds heat through the casing into passing air, which means its cooling depends on road speed — exactly the thing that is lowest when you most need power.

On a long climb the sequence is predictable. The motor draws high current at low rpm, where it is least efficient, so a large share of the energy becomes heat rather than motion. The windings warm, then the magnets. As temperature rises the motor becomes less efficient still, which produces more heat. Most controllers respond by reducing current to protect the hardware.

From the saddle this feels like the bike quietly giving up. Speed bleeds away, the assist becomes vague, and pedalling harder does not bring it back. It usually recovers after a few minutes of easy riding, which is why the problem is often misdiagnosed as a battery fault.

Sustained climbing is, more than anything else, what cooks a hub motor. Seventeen of the twenty-two bikes here are not mid-drives, and two of those — the AMYET S8 and the AMYET Ares 2 — drive both wheels, though AMYET never says where either motor sits.

What reduces it

Keeping road speed up, for airflow and for a motor turning nearer its efficient range.
Contributing power yourself. Every watt you add is a watt the motor does not have to make.
Splitting the load. Two motors sharing a climb each run cooler than one doing all of it — the quieter argument for AMYET's dual-motor S8 and Ares, and for the dual-hub AWD option on EUNORAU's FLASH Lite.
Not stopping halfway up. Restarting on a steep gradient means the highest current at the lowest speed, which is the worst combination for heat.

A mid-drive is not immune, but it is better placed. It sits in open air at the cranks rather than sealed inside a rim, and because the gears let it spin fast while the wheel turns slowly, it can stay near its efficient range on a gradient that would have a hub motor labouring.

Hub versus mid-drive, and why gearing changes everything

A mid-drive motor sits at the cranks and delivers its torque through the chain and the cassette. When you drop into a low gear, the drivetrain multiplies the motor's torque exactly as it multiplies yours. A mid-drive in its lowest gear is a different machine from the same mid-drive in its highest.

A hub motor sits in the wheel, downstream of the gears. Whatever torque it makes arrives at the wheel unmultiplied, and changing gear does nothing to it. A 7-speed hub-motor bike has one motor gear, and the manufacturer chose it for a sensible cruising speed rather than for a 12% gradient.

This is why the five mid-drives on this site are in a different category from the seventeen bikes that are not mid-drives, and why the 160 Nm SPECTER-S is not merely "twice as good" as an 85 Nm SF20H. The 85 Nm arrives at the wheel as 85 Nm. The 160 Nm arrives at the chainring and is then multiplied by whatever gear ratio you have selected.

The five are worth naming, because they are no longer one brand's speciality: EUNORAU's FLASH, SPECTER-S and SPECTER-ST, VTUVIA's FMB and Velotric's Discover M, with EUNORAU's FLASH Lite offering a mid-drive as one of three motor-position options. Velotric's is the least powerful of them on paper — 500W rated, 100 Nm — and the best geared, through a 9-speed 11–46T. On a gradient, gearing is the half of that pair that does the work.

11 speeds

SRAM NX on the SPECTER-S and SPECTER-ST, the only 11-speeds here

Gear count, and why it is a climbing specification

Count the gears on the twenty-two bikes and the pattern is stark. Seven speeds on the AMYET G60, EB26, S8, K10 and Ares; seven on the VTUVIA SF20H, Zeal LT7, SX20 and SN100; seven on the EUNORAU META26. Eight — Shimano — on the Velotric T1 ST Plus, Nomad 1 Plus, Packer 1, Discover 2 and Discover 3. Eight on the EUNORAU FLASH. Nine — Shimano Cues 11–46T, the widest ratio spread here — on the mid-drive Velotric Discover M. Eleven — SRAM NX — on the EUNORAU SPECTER-S and SPECTER-ST. One speed, with a 36T chainwheel, on the EUNORAU FLASH Lite. And nothing published at all on the VTUVIA Gemini or the VTUVIA FMB, the second of which is a $2,999 mid-drive whose gearing decides what its 160 Nm is worth.

On the seventeen bikes that are not mid-drives, gear count governs one thing: your own contribution. That is not nothing. Spinning a low gear at a comfortable cadence for the length of a climb can add a sustained couple of hundred watts from a moderately fit rider — power that goes straight into reducing motor load, and therefore motor heat, and therefore how much cutback you feel near the top. A geared bike also remains a bicycle when the pack is empty, which on a hilly route is not an academic point.

On the five mid-drives, gear count governs the motor as well as you. A wide-range cassette is what converts a torque figure into climbing ability, which is why EUNORAU's own listings pair the SRAM NX 11-speed with the 160 Nm of both SPECTER models, why Velotric pairs its widest 11–46T cassette with the Discover M's mid-drive rather than with any of its hub bikes, and why the FLASH's eight Shimano gears matter more than the same eight gears do on a Velotric hub bike. The trade is wear: all that force runs through your chain and cassette, which wear faster on a mid-drive than on a hub-drive bike.

Two motors: a traction answer as much as a power answer

Dual-motor bikes are sold on power, but on a hill their more important contribution is often grip.

As a bike climbs, weight transfers rearward. That helps a rear hub find traction, up to the point where the front wheel goes light and begins to wander. On loose ground — gravel, wet leaves, sand, snow — a single rear hub applying high torque simply spins, and no amount of extra wattage fixes a wheel that has already broken traction.

A second driven wheel changes that. The same total torque is split across two contact patches, so each is asked for less, and the front wheel pulls rather than merely steering. It also halves the thermal load per motor on a sustained climb — a quieter benefit than the acceleration figures, and a more useful one.

The AMYET S8 is the bike here that states all-wheel drive outright, with dual 1000W motors — position unstated — against the single rear hubs AMYET does name on the G60 and EB26. It is no longer alone in having two driven wheels: AMYET's folding Ares carries the same dual 1000W arrangement without the AWD wording or a slope claim, and EUNORAU's FLASH Lite offers a dual-hub AWD configuration at 52V 1500W with a published 184 Nm combined — the only two-motor arrangement here with a torque figure attached. At $1,049 against $689 and $669 the S8 costs money, weight and consumption, and AMYET still publishes no torque figure for either of its motors, or for the Ares's. Our single vs dual motor guide and the G60 vs S8 comparison both work through the trade.

Reading a maximum slope claim

Gradient is quoted two ways and the difference is large enough to mislead. A percentage is rise over run: a 10% grade climbs one metre in every ten travelled horizontally. Degrees measure the angle itself. The two are related by a tangent, and they diverge sharply as the slope steepens.

10% grade is about 5.7 degrees.
20% grade is about 11.3 degrees.
30% grade is about 16.7 degrees.
35 degrees is a grade of about 70%.

For context, the steepest paved public roads tend to fall in the 20 to 30% region, and a climb most riders would call brutal is usually between 12 and 18%.

Treat any maximum-slope number as a rough ranking signal between models in the same range rather than as something you can plan a route around. It tells you that AMYET positions the S8 above its single-motor bikes for gradient. It does not tell you how the bike behaves three minutes into a real climb with a rider and a rucksack on it. A published torque figure would.

What goes up: braking is half the problem

Buyers shopping for hills think about climbing. The descent is the part that actually demands hardware.

A heavy e-bike descending a long gradient turns a great deal of potential energy into heat in the brakes, continuously, for as long as the hill lasts. Once the pad compound overheats, braking force falls away — fade. On cable-operated brakes your hands are doing the work, so hand fatigue on a ten-minute descent is a real limit rather than a theoretical one.

Mechanical versus hydraulic, and who publishes a rotor size

Mechanical discs pull the pad through a cable. Lever feel is softer, force depends on hand strength, and cables stretch and need adjustment as pads bed in. Hydraulic discs move fluid instead: they need far less hand force, self-adjust for pad wear, and modulate more finely, which matters most when you are tired and the surface is loose.

Across the twenty-two bikes, eighteen have a hydraulic claim. AMYET lists hydraulic discs on the S8 and the K10, mechanical discs on the EB26, and describes the G60 only as dual disc and the Ares only as disc brakes, without a hydraulic claim on either — which normally means mechanical, since brands that fit hydraulics say so. Every EUNORAU bike is listed with hydraulic discs, as is every VTUVIA bar the FMB, for which VTUVIA publishes no brake specification at all — a conspicuous silence on a $2,999 full-suspension bike meant for descending. The SF20H, Zeal LT7, SX20, META26, FLASH and SPECTER-S state a 180mm rotor with it; the Gemini, SN100, FLASH Lite and SPECTER-ST state hydraulic discs without a rotor size. Every Velotric is hydraulic, and five of the six name TEKTRO: 180mm front and rear on the Packer 1 — which goes further than anything else here with a four-piston caliper, where every other caliper on this site is a two-piston — and 180mm front and rear on all three Discover models, each with a motor power cut-off in the levers. Velotric lists TEKTRO hydraulic discs on 160mm rotors for the T1 ST Plus, the smallest rotor here and a sensible match for a 39 lb bike, and hydraulic discs on the Nomad 1 Plus without stating a rotor size. VTUVIA goes further on the SF20H, naming LOGAN calipers and martensitic stainless rotors. EUNORAU goes further again on the META26 and the SPECTER-S, which have motor cut-off switches in both brake levers — power dies the instant you reach for a lever, which on a throttle bike descending a hill is worth having.

AMYET publishes no rotor diameter for any of its five bikes. Rotor size is a major determinant of both outright stopping power and heat capacity, and on bikes carrying this much mass that omission matters. If your riding involves repeated long descents, treat braking as the first specification rather than the last — see the EB26 vs S8 comparison.

Weight, and planning range for a hilly route

Mass works against you in both directions: it is what the motor must lift on the way up and what the brakes must control on the way down.

Eleven of the twenty-two bikes publish a weight. Velotric's T1 ST Plus is the lightest here at 39 lb, well clear of the VTUVIA SF20H at 60 lb, followed by the Discover 3 at 61 lb, the Zeal LT7 at 62 lb, the META26 at 68.4 lb, the SX20 and the step-thru Nomad 1 Plus at 70 lb, the high-step Nomad at 71 lb and the Packer 1 at 75 lb — the heaviest hub-motor bike here. Both EUNORAU flagships — the FLASH and the SPECTER-S — are listed at 50 kg, approximately 110 lb, nearly three times the T1. AMYET publishes no bare bike weight for any of its five models; the EB26 listing shows 30 kg, roughly 66 lb, but that is a shipping figure including packaging rather than the weight of the bike. VTUVIA publishes none for the Gemini, the SN100 or the FMB, EUNORAU none for the FLASH Lite or the SPECTER-ST, and Velotric none for the Discover 2.

On flat ground the difference is barely noticeable. On a 12% climb it is the whole story — and on the descent that follows, it is the brakes' problem.

Budgeting energy for elevation

Where a heavy fat-tire e-bike might use roughly 25 watt-hours per mile on mixed flat riding, sustained climbing can push that beyond 40, and a dual-motor bike with both wheels driving higher again.

÷ 40Wh

planning figure for a hilly route

Take each published capacity and divide by 40 watt-hours per mile. These are calculations from merchant figures, not measurements, and this site has not ridden any of these bikes.

Velotric T1 ST Plus — 352.8Wh ÷ 40 = about 9 miles. It is much the lightest bike here on much the narrowest tires, so it will do better on a climb than the arithmetic suggests — but it starts from far less energy than anything else here.
VTUVIA Zeal LT7 — 624Wh ÷ 40 = about 16 miles.
VTUVIA SF20H, SX20, SN100 and FMB — 672Wh ÷ 40 = about 17 miles each.
Velotric Nomad 1 Plus — 691Wh ÷ 40 = about 17 miles.
Velotric Packer 1 — 691.2Wh ÷ 40 = about 17 miles, and less again with a loaded rack.
Velotric Discover 2 — 705.6Wh ÷ 40 = about 18 miles.
AMYET EB26 and AMYET K10 — 720Wh ÷ 40 = 18 miles each.
EUNORAU META26 — 720Wh ÷ 40 = 18 miles on one pack; 1,440Wh ÷ 40 = 36 miles with the optional second battery.
Velotric Discover 3 — 730Wh ÷ 40 = about 18 miles.
Velotric Discover M — 801.6Wh ÷ 40 = about 20 miles, and of the Velotrics this is the one whose mid-drive and 11–46T cassette will beat the assumption on a gradient.
EUNORAU FLASH and FLASH Lite — 832Wh ÷ 40 = about 21 miles on the standard pack; roughly 2,808Wh ÷ 40 = about 70 miles with all three fitted.
EUNORAU SPECTER-S — 840Wh ÷ 40 = 21 miles; 1,560Wh ÷ 40 = 39 miles with the second battery.
EUNORAU SPECTER-ST — 840Wh ÷ 40 = 21 miles; roughly 1,656Wh ÷ 40 = 41 miles with the second battery.
AMYET V9-G60 — 960Wh ÷ 40 = 24 miles.
AMYET Ares 2 — 1,040Wh ÷ 40 = 26 miles, and running both motors on loose ground will push consumption higher still.
AMYET S8 — 1,200Wh ÷ 40 = 30 miles, with the same caveat.
VTUVIA Gemini — no figure. VTUVIA describes the battery as "dual 52V 21Ah" without saying whether that is one pack or two, so the division cannot be done.

Notice how far those sit below the headline claims — AMYET lists 30–65 miles for the G60, up to 60 for the EB26, 35–60 for the K10, 70–75 for the S8 and nothing at all for the Ares; VTUVIA lists up to 70 for the SF20H, SX20, Zeal LT7 and SN100, up to 90 for the Gemini and a more modest 30–50 for the FMB; EUNORAU lists up to 100 for the META26 on two packs, up to 220 for both the FLASH and the FLASH Lite on three, and 80 for the SPECTER-ST on two; Velotric lists 70 for the T1 ST Plus, 55 for the Nomad 1 Plus, 52 for the Packer 1, and 75, 80 and 95 for the Discover 2, Discover 3 and Discover M — and is the only brand here that tells you the conditions behind those figures, a 187-lb rider, on PAS 1, on flat roads. Nothing is wrong with the published numbers as ceilings. They describe gentle riding on flat ground, which is not what a hill route is.

Plan against the lower figure, read our range guide for the full arithmetic, and if you are also close to a payload ceiling, the guide for heavier riders covers how mass interacts with wheels and brakes. The full buying framework puts terrain in order against everything else.

Common questions

How many watts do I need to climb hills on an e-bike?

Wattage is the wrong question, though a higher rated figure does help. Watts describe electrical input; newton-metres describe how hard the motor turns the wheel, and that is what a gradient demands. A 350W Velotric T1 ST Plus publishes 40 Nm, a 500W EUNORAU META26 publishes 55 Nm and a 750W VTUVIA SF20H publishes 85 Nm, so within hub motors the more powerful bike is also the stronger climber — but a 1000W hub motor with no published torque figure cannot be placed on that ladder at all. Compare rated power rather than peak, look for newton-metres, and treat gearing and your own fitness as part of the system.

Is a mid-drive motor really better for hills than a hub motor?

For steep, sustained climbing, generally yes, and the reason is gearing rather than the torque figure alone. A mid-drive turns the chainring, so dropping into a low gear multiplies the motor output exactly as it multiplies yours. A hub motor sits downstream of the gears, so its torque arrives at the wheel unchanged whatever you select. EUNORAU publishes 160 Nm for the SPECTER-S and SPECTER-ST mid-drives through SRAM NX 11-speed gearing and 220 Nm for the FLASH mid-drive through eight Shimano gears, VTUVIA publishes 160 Nm for the FMB, and Velotric publishes 100 Nm for the Discover M through a 9-speed 11–46T — against a maximum of 100 Nm from any single non-mid-drive motor we cover, on the VTUVIA Gemini, and 92 Nm from the strongest confirmed single hub, EUNORAU’s FLASH Lite — neither with gears to multiply it. The trade is chain and cassette wear.

Why does my e-bike lose power on long climbs?

Almost always heat. A hub motor drawing high current at low road speed is working at its least efficient point, and it cools mainly through airflow that is barely present when you are grinding uphill. As the windings warm, the controller reduces current to protect the motor, which feels like the bike quietly giving up. It usually recovers after a few minutes of easier riding, which is why it is often mistaken for a battery fault. Keeping road speed higher, pedalling in a low gear and avoiding restarts mid-climb all reduce how often it happens.

Which of the twenty-two bikes publishes the most climbing torque?

The EUNORAU FLASH 2.0, at 220 Nm from a 1000W Truckrun mid-drive, delivered through an 8-speed Shimano cassette that multiplies it further in the low gears. Three bikes follow at 160 Nm, all from a Bafang G510 mid-drive: the EUNORAU SPECTER-S 3.0 and SPECTER-ST 2.0, both through SRAM NX 11-speed, and the VTUVIA FMB, which publishes no drivetrain at all. Velotric’s Discover M states 100 Nm from a mid-drive through a 9-speed 11–46T. Outside the mid-drives the VTUVIA Gemini leads at 100 Nm, from a motor VTUVIA does not place, ahead of the EUNORAU FLASH Lite’s rear hub at 92 Nm, the VTUVIA SF20H and SX20 at 85 Nm, the Velotric Nomad 1 Plus, Packer 1, Discover 2 and Discover 3 at 75 Nm, the Zeal LT7 at 65 Nm, the EUNORAU META26 at 55 Nm and the Velotric T1 ST Plus at 40 Nm, which is the lowest figure here. AMYET publishes no torque figure for any of its five bikes and VTUVIA publishes none for the SN100, so those six cannot be ranked on this measure at all.

What does a 35-degree maximum slope claim actually mean?

Less than it sounds. Thirty-five degrees converts to a gradient of roughly 70%, far steeper than any paved public road in normal use — the steepest tend to fall in the 20 to 30% range. Figures like this are typically produced from a short burst on a test ramp with a light rider rather than from sustained climbing. AMYET markets that figure for the S8. Treat a maximum-slope claim as a rough ranking signal between models from the same brand, not as a specification you can plan a route around, and prefer a published torque figure where one exists.

Do I need hydraulic brakes for hilly riding?

On a heavy e-bike ridden on long descents, they are the sensible choice rather than a luxury. Hydraulic discs need much less hand force, self-adjust as pads wear, and modulate more finely when the surface is loose. Eighteen of the twenty-two bikes here carry a hydraulic claim. The VTUVIA SF20H, Zeal LT7 and SX20 and the EUNORAU META26, FLASH and SPECTER-S are listed with hydraulic discs and 180mm rotors, as are Velotric’s Packer 1 — which uses a four-piston caliper where everything else here is a two-piston — and all three Discover models, on TEKTRO 180mm front and rear with a motor power cut-off in the levers. Velotric lists TEKTRO hydraulic discs on 160mm rotors for the T1 ST Plus and hydraulic discs without a rotor size for the Nomad 1 Plus, as VTUVIA does for the Gemini and SN100 and EUNORAU for the FLASH Lite and SPECTER-ST. The META26 and SPECTER-S add motor cut-off switches in the levers. AMYET lists hydraulic discs on the S8 and the K10, mechanical discs on the EB26, unqualified dual disc on the G60 and unqualified disc brakes on the Ares — and publishes no rotor diameter for any of the five. VTUVIA publishes no brake specification at all for the $2,999 FMB.

How much range do hills cost?

A great deal. Where a heavy fat-tire e-bike might use around 25 watt-hours per mile across mixed flat riding, sustained climbing can push consumption past 40, and running two motors on loose ground pushes it higher still. As a planning estimate, divide the pack capacity by 40 rather than 25 for a hilly route. That gives about 16 miles for the Zeal LT7 on 624Wh, 24 miles for the AMYET G60 on 960Wh, and 30 miles for the S8 on 1,200Wh — all well below the headline claims, which describe flat ground and gentle assist.

Keep reading