Most e-bike OEM buyers do not choose between a hub motor and a mid-drive because of marketing. They choose because a retail price point, a target market and a test ride have already narrowed the options, and the last decision is which architecture to put in the BOM. The two drive layouts are not interchangeable parts of equal capability at different prices. They fail in different ways, they suit different frame designs, and they carry different service liabilities over the life of the fleet.
This guide is written from the sourcing side of that decision. It sets out how each architecture actually works, what the real efficiency and torque numbers look like, where the cost delta sits in the bill of materials, and which product categories in our own catalogue map onto which drive type. Every specification below comes from the motor and controller platform we build, a 250 W to 1500 W range across geared hub, direct-drive hub and mid-drive units at 36 V to 72 V, not from a generic industry summary.
How the Two Architectures Work
A hub motor is built into the wheel. The stator and windings sit in a housing laced to the spokes, and the motor drives the wheel directly or through an internal planetary gear set. Because it drives the wheel rather than the cranks, it is mechanically independent of the bicycle's gearing. The rider can be in any gear, or not pedalling at all on a throttle-equipped model, and the motor still delivers its rated output to the road.
A mid-drive motor sits at the bottom bracket and drives the crank axle. Motor torque passes through the chainring, the chain and the rear cassette before reaching the wheel. That single mechanical difference is the source of almost every trade-off in this comparison: the mid-drive can use the bicycle's gears to multiply its torque, and it can also overload them.
| Parameter | Geared hub | Direct-drive hub | Mid-drive |
|---|---|---|---|
| Power range we build | 250–1500 W | 500–3000 W | 250–1000 W |
| System voltage | 36 / 48 / 52 / 60 / 72 V | 48 / 60 / 72 V | 36 / 48 V |
| Peak efficiency zone | 80–92% | 80–90% | 80–92% |
| Motor mass at the axle | 3.5–5 kg rear | 6–9 kg rear | 3–4 kg at bottom bracket |
| Drivetrain wear | Pedal load only | Pedal load only | Pedal plus full motor torque |
| Regenerative braking | Not practical (freewheel) | Yes | Yes on some units |
| Controller interface | Sine wave FOC | Sine wave FOC | Sine wave FOC, torque sensor input |
| Motor ingress rating | IP65 | IP65 | IP65 |
One specification detail matters more than buyers expect: the controllers in all three families are sine wave field-oriented control units, potted to IP54 to IP65, with regenerative braking, cruise control, walk mode and PAS sensor input available, and UART, CAN bus or Bluetooth telemetry output. The controller choice is not what separates a hub line from a mid-drive line. The mechanical interface to the bike is. That is also why motors and controllers are supplied as matched pairs and dyno-tested together at five load points before shipment, with the test report included in every production batch.
Torque, Climbing, and Gear Multiplication
The torque argument is the one most often stated badly in both directions. A hub motor's torque at the wheel is fixed by its winding and its controller current limit. It cannot be multiplied. If a 500 W geared hub produces a given wheel torque, it produces that same torque whether the rider is in first gear or top gear, at walking pace or at 25 km/h. Climbing ability is therefore a function of motor spec alone.
A mid-drive's torque at the wheel is the motor's output multiplied by whatever gear ratio the rider selects. In a low gear on a 1:3.5 ratio, the same motor output reaches the wheel multiplied several times over, which is why a 250 W mid-drive can walk a loaded bike up a grade that stalls a 750 W hub. The catch is that this multiplication is limited by what the chain and cassette can survive, and by the fact that the rider must actually shift down.
| Duty condition | Geared hub 500 W | Mid-drive 250 W |
|---|---|---|
| Flat, 20 km/h cruise | Adequate, low motor load | Adequate, low motor load |
| 8% grade, 100 kg system | Motor near current limit, heat builds | Comfortable in a low gear |
| 8% grade, 180 kg cargo system | Thermal derate likely within minutes | Sustainable if gearing is low enough |
| Sustained steep climbing | Not recommended | Design case |
The practical consequence is a market split that shows up in our order book. The EU market, where EN 15194 caps continuous rated power at 250 W, is dominated by mid-drive systems for any city with real topography. A 250 W hub motor cannot climb; a 250 W mid-drive can. In the US, where class limits run to 750 W, hub motors dominate because the extra power compensates for the lack of gearing. Our motor power guide covers the power-class rules in more detail.
Efficiency and Real-World Range
Both architectures peak in the same 80 to 92 percent band, so any claim that one is simply more efficient is wrong. The difference is where each one operates relative to its peak, and that is decided by the route.
- Flat, steady cruising: hub and mid-drive are within a couple of percentage points. A direct-drive hub can be marginally better because it removes the planetary gear stage.
- Stop-start city riding: a mid-drive keeps the motor near its efficient band through every acceleration because the rider can downshift; a hub motor spends more time at low RPM and high current.
- Hilly terrain: the mid-drive's advantage is largest here, because gear multiplication lets the motor stay in its efficient RPM range instead of stalling into a high-current, low-speed condition.
- Heavy cargo: a hub motor under sustained low-speed load runs hot and thermally derates; a mid-drive distributes that load across the drivetrain instead of concentrating it in the hub shell.
For a buyer projecting fleet range, the efficiency delta becomes a battery purchasing decision. If a mid-drive cargo line can meet a duty cycle on a smaller pack than a hub line equivalent, the battery saving offsets part of the motor premium. Battery chemistry, cell grade and pack sizing economics are covered in our battery technology guide, and the pack-level selection logic in the battery pack OEM guide.
Weight Distribution and Handling
A geared hub motor puts roughly 3.5 to 5 kg at the rear axle, and a direct-drive unit puts 6 to 9 kg there. That mass improves rear traction, which is genuinely useful on fat-tire, sand and snow builds, but it makes the bike rear-heavy to lift, slower to change direction, and harder on the rear wheel and rack. In cargo applications it also competes directly with payload on the rear wheel's load rating.
A mid-drive puts 3 to 4 kg low and central at the bottom bracket, which is the best place on a two-wheeler for mass, and it keeps both wheels light. Suspension response improves because unsprung mass at the rear drops, and the bike feels more balanced when loaded. For commuter and mountain lines this is a clear handling argument. For fat-tire builds the traction trade-off is real, which is why many fat-tire programmes accept the hub's handling penalty; our fat-tire e-bike OEM guide covers that category specifically.
Maintenance and Total Cost of Ownership
The maintenance profiles diverge after the first few thousand kilometres, and this is where fleet economics are decided. A geared hub motor is a sealed unit: the planetary gears run in grease, the bearings are the wear item, and the chain and cassette carry pedal load only, so they last substantially longer than on a mid-drive. For rental, sharing and delivery fleets this is usually the deciding factor, because motor service is a bearing replacement rather than a drivetrain rebuild, and it can be done without touching the transmission.
A mid-drive sends full motor torque through the chain, cassette and chainring. Drivetrain replacement intervals shorten accordingly, and mid-drive chains are typically e-bike-rated parts with reinforced pins rather than standard bicycle chain. The motor itself is serviceable, our mid-drive units open for gear and seal replacement, but it needs a trained technician and the bike is out of service while it happens. Fleet operators should budget additional drivetrain parts per bike per year on a mid-drive line, and should plan the service interval into the contract rather than discovering it in year two.
Two maintenance details are specific to our platform and worth designing around. First, cable exit position is a configurable option at the axle side, which lets you route the motor cable inside the chainstay on frames that allow it and remove a snag point on fleet bikes. Second, controller potting grade is selectable between IP54 and IP65, so a pressure-wash-heavy depot duty can be specified up rather than inheriting a commuter-grade default.
OEM Cost Comparison: The BOM Impact
For an OEM buyer the cost delta is the number that decides the programme. The gap is not only the motor: a mid-drive changes the frame interface and the drivetrain specification at the same time.
- Motor and controller: a mid-drive unit carries a higher unit cost than a geared hub of the same power class, before any drivetrain changes.
- Drivetrain: a mid-drive build needs an e-bike-rated chain, a cassette rated for motor torque, and typically a stronger chainring, all of which add cost per unit and add a wear item to the service schedule.
- Frame: a mid-drive frame needs a reinforced bottom bracket shell and chainstay area to take the torque input, which restricts which open-mould frames you can use and adds tooling or modification cost.
- Hub frame: a hub build needs the dropout width and axle standard matched to the motor, an M12 or M14 axle with flat, which is a simpler and cheaper constraint to design around.
- Sensor set: a mid-drive line usually specifies a torque sensor for ride quality; a hub line can run a cadence sensor and PAS input at lower cost.
The consequence is that the two architectures rarely compete head-to-head in the same retail band. Below roughly $1,300 retail, a mid-drive's added motor, drivetrain and frame cost cannot be absorbed, and hub motors own the segment. Above about $1,800 retail, the margin carries the mid-drive premium and the climbing and handling benefits become the selling point. Between those two numbers the decision is genuinely market-specific, and it usually turns on whether the target city has hills. Full procurement terms, MOQ structure and lead times are in our OEM e-bike buying guide, and the component-level selection logic for the controller side is in the motor controller guide.
Decision Framework by Product Category
The architecture maps cleanly onto product categories. Use this as a starting point and then adjust for the specific market and terrain:
- City commuter, flat terrain: geared hub 350–500 W. Lowest cost, lowest maintenance, sufficient performance, and the simplest frame interface.
- City commuter, hilly terrain or EU 250 W limit: mid-drive 250 W. The only way to get real climbing performance inside the EN 15194 continuous power cap.
- Fat-tire, snow and sand: geared hub 500–750 W. Rear traction is a benefit rather than a penalty, and the simpler drivetrain survives grit better.
- Mountain and trail: mid-drive. Central mass, gear multiplication and torque sensor response all matter more off-road than the cost delta.
- Cargo and heavy payload: mid-drive where gradients are real; hub acceptable on flat urban routes where payload stays within the wheel rating.
- Fleet, rental and sharing: geared hub. Motor service intervals and drivetrain wear dominate the total cost of ownership at fleet scale.
Two further variables should be settled at the same time as the architecture, because they determine the compliance package. The power class you choose, 250 W for the EU, 750 W for the US class system, or higher for moped registration, decides which certification and registration regime applies. And the magnet grade, N38SH or N40H in our range, decides how much thermal headroom the motor has in a hot climate or a sustained-climb duty. Certification is covered in our EU vs US certification guide; the compliance documents, CE, RoHS and the EN 15194 system-level test package, are compiled per order.
Both architectures are built on the same platform in our factory, with geared hub systems from 250 W to 1500 W for commuter, fat-tire and fleet lines, and mid-drive builds for premium commuter, mountain and cargo programmes. Orders start at 500 units for motors and 200 units for controllers per configuration, with a 20 to 30 day standard lead time and pre-production samples shipped before mass production. Every motor and controller pair is dyno-tested at five load points before it leaves, and the test report travels with the batch.
Get a Motor Architecture Recommendation for Your Line
Tell us your target market, terrain, retail price point, and duty cycle. We will spec the motor, controller, and drivetrain, with cost breakdown and certification documents for your region.
