Is DJI Reshaping the Bike Industry?
Avinox and Amflow may be disrupting the mountain bike industry with implications far beyond anything we've seen for decades (or ever)...
This week, was a wild one in the bike industry. In roughly 48 hours at least 20 companies all announced new electric mountain bikes. For those who don’t track the space, I haven’t seen this many new bikes announced in such a small of a time window since the Interbike era two decades ago. What spurred the mass-launch from brands around the globe was the official announcement of a new drive unit from drone and camera maker DJI. According to the press release, sixty brands have now committed to the Avinox (DJI’s drive unit subsidiary) platform, which is incredible considering the company hadn’t even touched industry 3 short years ago.
Change can be hard to spot in any industry. Apple had just 3% of the personal computer market back in the 90s but now moves roughly $300B worth of hardware per year. Kodak dominated the camera world in the 80s but is bankrupt today. Technological change often result in foundational shifts within a competitive landscape.
That’s what prompted this piece. Not the motor specs (though we’ll get there). Not the geopolitics (though we’ll touch on that too). The sheer velocity of adoption and what it might signal for the broader landscape.
Is DJI fundamentally reshaping the bike industry as we know it?
Let’s dig in.
Disclaimer: This post is not about trail access and what more powerful motors might mean for the sport. There are plenty of reddit forum posts and blogs addressing this issue, and my hot take isn’t going to sway the public’s view one way or another.
What Is a Drive Unit, Really?
Before we talk about Avinox, we need to talk about what a drive unit actually is on a modern eBike, because it’s not a “component” the way a derailleur or a set of brakes is a component. It’s something fundamentally new to the bicycle world, and I think most people (including many in the industry) haven’t fully processed the impact it has on a complete bicycle.
A drive system includes: the motor, the battery, the battery management system, the sensors (dozens of them), some kind of display, output controls, a companion app, firmware, the charging system, and orchastration software layer that governs all of it.
No one in the bicycle industry has ever had a single supplier control this much of a bike’s identity. Not only do you have to build your frame around the motor and battery specs, but the supplier of these components largely dictate how the bike performs in a way that a consumer feels far more than seemingly any other part on the bike (outside maybe suspension).
Put simply, one might argue the most important part of an e bike is the drive unit, and there is no interchangeability like there might be of other parts on the bike. Once you buy in, you are sticking with it at least for as long as you sell that particular frame. Considering the pace of change has dropped dramatically with respect to frame iterations, this should be a big deal for those paying attention.
The Specs (and Why the Gap Is So Large)
Okay, let’s talk numbers, because clearly there is a reason the industry is going toward Avinox. Here’s where the M2S (launched this week) sits relative to the competition:
The M2S produces 2.5x the peak power of a Bosch CX at roughly the same weight. The hypothetical support ratio is double. On paper this motor is significantly better than really anything else out there.
I haven’t ridden the new motor yet (any brands reading this, reach out), but by according to all the online media outlets, the new motor is absolutely phenomenal. Two of my friends talk of the M1 (prior generation motor) in a way that reminds me of early iPhone adopters raving of their new device compared to incumbent blackberry devices - the entire experience is just plain better.
As to real world comparative testing, I haven’t seen anything quantitative with the new drive units. However, German test site emtb-test.com ran a standardized climb test across all major motors including the prior gen M1. Their finding: the (older) required only 170 watts of rider input to complete the test. Bosch CX required 316 watts. Yamaha (nobody uses this anymore) required nearly 380 watts. The Avinox motor is so powerful and efficient that the rider input required is cut nearly in half compared to Bosch.
Turning back to the just released drive units, the M2S just blew the existing gap even wider. It produces 50% more power than the M1 while weighing only 110 grams more. Avinox also replaced the high-resistance oil seal with sealed bearings, reducing unassisted pedaling resistance by 41%. They switched from straight-cut to helical gears, making it quieter. They added a 700Wh battery option, removable battery options, and a 7A fast charger that goes 0-80% in just over an hour.
Every objection the industry had about the M1 (”it’s draggy when the motor is off,” “no removable battery,” “the rattling noise”), they fixed. While simultaneously making it 50% more powerful.
Apparently this is what happens when a company with 14,000 engineers and a functionally unlimited R&D budget decides to iterate fast. Culture matters and if their drones are any indication, DJI might have something special under the hood when it comes to shipping incredible hardware+software, fast.
So Who Is DJI?
Most people in the bike world know DJI as “the drone company.” That framing is like calling Apple “the computer company” in 2007. Technically true. Wildly incomplete.
DJI (SZ DJI Technology Co., Ltd.) was founded in 2006 by Frank Wang, born in 1980 in Hangzhou. Wang was, by his own admission, a mediocre student. He dropped out of East China Normal University, transferred to Hong Kong University of Science and Technology, and caught the attention of a robotics professor through a helicopter flight control system project. He moved to Shenzhen with two classmates, renting a tiny space with about $28,000 from his parents.
The early culture was, to put it charitably, intense. Wang is a notorious perfectionist who would call employees at any hour, gave excruciatingly detailed instructions (down to how to tighten each individual screw), and churned through staff at an alarming rate. His co-founder sold his apartment to fund the company. Another early employee paid a $4,600 contract-breaking penalty to join, earning the nickname “Thirty Thousand Chen.”
That culture produced the company that has since dominated the drone industry. DJI now employs roughly 14,000 people across 17 international offices. They control somewhere between 70-80% of the global consumer drone market. Revenue figures are hard to pin down (DJI is private), but market rumors put 2025 revenue approaching 80 billion yuan (roughly $11 billion), and Frank Wang has said internally he expects the company to stably exceed 100 billion yuan.
DJI’s core competencies are electric motors, battery management systems, sensor fusion, software, and miniaturization. That is the exact technology stack required to build the best eBike drive system. From a strategy standpoint, they probably looked at the eBike motor market and saw a $13 billion opportunity being served by companies (Bosch, Shimano) iterating on 3-year product cycles with the urgency of a German pension fund.
The internal culture that pits teams against each other, demands obsessive iteration, and ships product at startup speed? That’s how you go from zero to market-leading motor in under two years (or build drones the world had never seen 15 years ago). Bosch has been making eBike motors for over a decade. DJI beat them on every spec in their first attempt.
Apples and Oranges: The Balance Sheet
Let’s put DJI’s scale in context against the industry they entered.
The global mountain bike market, depending on whose estimate you trust, is somewhere between $2.5 billion and $14 billion (the range is wide because analysts can’t agree on what counts asa “mountain bike”). Let’s be generous and call it $10 billion.
DJI’s revenue is approximately 9.8-$11 billion with 2026 projections aiming at $14B.
One company that is likely larger than the entire global mountain bike industry. And that’s before you consider the broader eBike TAM, which is $55-70 billion globally and growing at 10-14% annually, at least according to Claude (take this with a grain of salt).
For additional context: Trek does roughly $1-2 billion in revenue. Giant maybe $2.5 billion across all categories. Specialized maybe $1-1.5 billion. DJI is bigger than all of them combined.
When Frank Wang’s internal communications talk about pushing past 100 billion yuan, the entire bicycle industry is (sort of) a side project. DJI could fund an Avinox loss for years without blinking. They could outspend every bike company’s R&D budget combined and it probably wouldn’t matter. Whats worse is this is a drone company with real free cash flow and positive earnings entering into an industry that is in shambles. While everyone else is on life support, they have the ability to go on the offensive unlikely literally any other company in the entire industry. Which brings us to the next part of the puzzle…
Amflow and the Vertical Integration Play
DJI launched Amflow in 2023 as the in-house bike brand (complete bikes) to showcase the Avinox system. “Just a bunch of tech experts with a limitless passion for riding mountain bikes,” per their own copy. In reality, Amflow was incubated by DJI as a platform to showcase and develop the motor system before spinning it off to third parties.
Hot take, but Amflow has become more than a demo vehicle. I have it on good authority that Amflow is Fox’s number one bike customer (or at minimum, their number one eBike customer). If this is true, its absolutely absurd and should be a blinking red light to the rest of the industry. A brand that didn’t exist three years ago is now moving more Fox suspension than companies that have been around for decades.
To add, they’re verticalizing fast. Here’s what Amflow already makes under their own brand:
Frames
Motor, battery, BMS, display, sensors, software (obviously)
Cranks (DJI Avinox SL Crank)
Carbon wheelset (HMC-30)
Alloy wheelset (HMA-30)
Handlebars and stem
What they’re still buying from western suppliers: Fox suspension, Schwalbe tires, SRAM drivetrains, Magura brakes etc.
I’m wagering the list of external suppliers will likely be getting shorter over time, not longer. Chinese technology companies have a well-documented tendency toward vertical integration. BYD, the Chinese EV giant, produces roughly 75% of its vehicle components in-house. Tires and windows are about the only things they outsource. DJI’s drone business follows the same pattern: they make their own motors, batteries, cameras, gimbals, flight controllers, and software. The entire stack.
The drivetrain is the obvious next target, and here’s why: current eBike drivetrains are a legacy architecture. Every chain, cassette, and derailleur on every eBike today was designed for human power output (maybe 200-400 watts sustained for a strong rider). The Avinox M2S produces 1,500 watts. That’s like bolting a Corvette engine onto a go-kart transmission. Chains stretch faster. Cassettes wear faster. Chainrings get chewed up. SRAM and Shimano have responded with beefier components, but nobody has fundamentally rethought the drivetrain for motor-assisted power delivery.
Pure speculation on my behalf, but DJI has the engineering depth to build a purpose-built eBike transmission. An internal gearbox integrated directly with the motor. A belt drive system. Maybe even a CVT optimized by software that reads terrain, cadence, and motor output in real time (which, if you squint, is basically a drone flight controller problem applied to a different domain). Something proprietary and better, especially for the mobility space.
Here’s a fun thought experiment: an Amflow with a proprietary drivetrain, belt drive, integrated motor+gearbox, near-zero drivetrain maintenance, and a $4,999 price point. I’d also wager they add a few new frames in different geometry and travel configurations.
If that doesn’t suit you, the high end boutique side of things probably will, with the Avinox selling complete drive units + drivetrains to bike brands offering longer travel steeds or those with exotic materials.
TL;DR - Amflow is aiming right at your value oriented high performance e mountain bikes. The bikes for the 80-90% of the market, and I don’t think this is a “fun way” they dabble in the industry, I think its a full fledged attack the same way a car company might offer their motors to other OEMs, but also sell the motor in their own packaging, often at the best pricing.
The Tariff Question (Brief, Because It’s Messy)
I’m not going to spend 2,000 words on this because I’ve written about tariffs before and frankly we’re all exhausted. But here’s what you need to know.
DJI’s drone business got effectively banned from the US market. The FCC added foreign-manufactured drones to its Covered List in December 2025, blocking new DJI models from receiving the equipment authorization needed for legal import and sale. Tariffs of up to 145% were reimposed on DJI drones.
In response, DJI restructured. The former “DJI Avinox” is now simply “Avinox.” The legal entity renamed from SZ DJI Technology Co., Ltd. to SZ Avinox Innovation Co., Ltd. According to Handelsblatt (the German equivalent of the Financial Times), DJI feared its eBike drive could fall victim to drone-related regulations, so it spun off the motor division.
This is more a slight of hand than anything real. Same people. Same technology. Same parent company. Same Shenzhen address. Different name on the door. Bikerumor described it accurately: “It is still the same tech developed by the original DJI-backed team, and it still shares DJI tech, like the same display.” E-Mountainbike Magazine confirmed: Avinox “remains a subsidiary of DJI.”
Whether this maneuver holds up long-term is an open question. DJI previously tried routing drone sales through shell companies in Malaysia before FCC component-level restrictions closed that door. Congress is already running the same playbook on Chinese humanoid robots. The regulatory risk is real.
On the tariff side specifically: e-bike lithium batteries from China face a 25% tariff in 2026, up from 7.5%. That’s significant. On the positive side, the White House just confirmed this week that bicycles, e-bikes, and frames will not be subject to new Section 232 tariffs on steel and aluminum, and some existing ones are being removed. PeopleForBikes led that fight with 1,300+ public comments.
For brands assembling bikes outside China (say, Taiwan) with a Chinese motor inside, the country-of-origin question is genuinely complex. US Customs applies a “substantial transformation” test. Building a complete bicycle from a motor, frame, suspension, drivetrain, wheels, brakes, and electronics is almost certainly a substantial transformation (the finished product has a different name, character, and use than any individual component). But CBP has been aggressive about scrutinizing these arrangements, and there’s legal precedent suggesting that when the “core component” (think: battery cells in a solar panel) originates in China, the whole product can be classified as Chinese-origin. Whether CBP views the motor as the “core component” of an eBike is a question that hasn’t been definitively answered.
Anyway, part of why I flag this is if the tariff thing changes at all, it could make Amflow and Avinox far more, or less, expensive depending on where you are buying from. And yeah, because DJI is in a weird spot with the US Govt.
The Big Picture
Let’s zoom out.
DJI isn’t just going after eMTB. Avinox-powered bikes are already showing up across urban, trekking, touring, gravel, and low-step categories. Velo de Ville (a German city/trekking brand) was one of the first urban adopters. Paprika 53 launched an Avinox-powered gravel bike. Velduro claims the first electric gravel bike driven by the system.
Mountain biking appears to be the beachhead. It’s where you prove the motor against the hardest use case and the most demanding customers. If your motor can handle 1,500W on a 150mm enduro bike being hammered down rock gardens at speed, it can handle a commuter on a bike path. You earn credibility in MTB, then you scale into the volume segments where Bosch actually makes its real money: city bikes, trekking bikes, cargo bikes. The millions of European commuters who currently ride Bosch-powered bikes aren’t loyal to Bosch. They’re riding Bosch because it’s on the bike they wanted. If Canyon (already an Avinox partner) puts an Avinox motor on a trekking bike that’s lighter, smarter, and $500 cheaper, those customers will switch when its time to buy a new commuter bike.
The global eBike market is $55-70 billion and growing at double digits. The eBike motor market specifically is roughly $13 billion. Avinox doesn’t need to “beat Bosch” to build an enormous business (though I think they can beat em). If they capture 10-15% of the global mid-drive motor market within five years, that’s a $1-2 billion revenue line growing at 15%+ annually. Layer on Amflow’s bike revenue and the margin capture from vertical integration, and you’re looking at a business that could rival the size of Trek or Specialized on its own.
I think the “can they take Bosch?” framing undersells what’s happening. Bosch’s lock-in used to be the moat, in a way. “Sure, the Avinox specs are better, but Bosch has the dealer network, the service infrastructure, the track record.” That argument sounds a lot like “sure, the iPhone is cool, but Nokia has the distribution” circa 2007…and we know what happened.
We’ve seen this movie before. We’ve seen what happens when a better technology company with deep engineering talent, massive scale, and a willingness to iterate at startup speed enters a market dominated by slower-moving incumbents. Sometimes the incumbents adapt. Sometimes they don’t. But the change is afoot.
I am, as the kids say, fully Avinox-pilled, even without having ridden one. I frankly wouldn’t even look at an e bike without that motor. For the first time in years, the market is genuinely exciting again, and we might see breakthroughs that have been unicornware for decades. Whether that’s good news or bad news depends entirely on which side of the disruption you’re sitting on.
If you’re a rider: enjoy the ride. Competition is about to make your next bike significantly better and possibly cheaper.
If you’re an executive at Bosch, Shimano, or SRAM: I’d start moving faster.
Editor's note: Since publishing, I've done additional research into the DJI/Amflow corporate structure. Chinese corporate registry data shows that Amflow is operated by Anflow Bicycle (Shenzhen) Co., Ltd., whose sole shareholder is a Hong Kong holding company called Worang Technology. The legal representative of the Amflow entity is Song Jianyu, a senior DJI management team member recruited personally by founder Frank Wang during the company's critical growth period. DJI has publicly described Amflow as 'incubated by DJI' and stated its intention to eventually let the brand operate independently, but there is no evidence of outside investors or independent financing to date. For all practical purposes, Amflow appears to have full access to DJI's balance sheet. Every incumbent eMTB brand should be modeling accordingly
Editor’s note X2: At Sea Otter I directly asked Avinox and Amflow about the relationship with DJI. They were adement there is no tie to DJI outside of being incubated by DJI. Avinox also suggested there is no tie with Amflow, and they are independent from one and other. This seems hard to believe considering they were sharing the same booth space. I will follow up on this as I learn more .
As always, thanks for reading. If you enjoy what I do, please subscribe and share. And if you like bikes, grassroots racing, and are looking for an alternative to Strava with features built for us, please sign up for my Rali App waitlist at www.raliapp.com - Oh and I’m serious about testing - I have a YouTube channel that does well and would love to throw a leg over one of these steeds!
Cheers, JB