Faraday Future has spent the better part of a decade promising to upend the car industry, and the results have landed somewhere between “modest” and “wait, is this company still going?”
The El Segundo outfit has trickled out only a small number of its flagship FF 91 since 2023, spun up a cheaper second brand called Faraday X, and then in May hit pause on its original 400-volt Super One program while it decides whether to go 800-volt electric or extended-range hybrid instead. So when word lands that FF has patented a new powertrain, the reflex is a tired sigh.
This one’s worth a second look. The U.S. Patent and Trademark Office just granted Faraday a transmission design built around one bet: pull the gasoline engine, the electric motor and the wheels apart from each other, so a controller can decide on the fly which one does what. The hardware is the patent. The “AI brain” meant to pick the mode is a separate system FF calls AIHER, and that’s where the company’s whole pivot is pointed.
The patent pulls the engine, motor and wheels apart
The grant is U.S. Patent No. 12,630,004, titled “Range-Extending Hybrid Transmission System,” issued in May 2026 from an application filed last June, with 13 claims. It went to a Faraday subsidiary saddled with the mouthful name Future AIHER.
The mechanical idea is in how the pieces are linked, or rather how they’re deliberately not. Per Faraday’s own filing summary, the system bundles a first rotary power source, a generator-motor, a differential and the driving wheels, strung along a set of shafts and clutches that can engage or disengage each part independently.
Most hybrids bolt the engine, motor and wheels together with a tightly choreographed run of gears and clutches. Faraday’s design lets them work concurrently or on their own. The engine can act purely as a generator, drive the wheels directly, do both, or shut off entirely while the motor handles things.
Faraday’s framing for the payoff is strong range extension with what it calls “weak hybridization” — long driving range without the full mechanical burden of a conventional hybrid. The layout is multi-shaft and multi-clutch, but the company argues each component carries fewer hard obligations to the others, so the routing logic ends up simpler.
One thing worth nailing down: the patent itself is the decoupled hardware. The AI part is AIHER (AI Hybrid Extended-Range), the range-extender platform Faraday set up as a dedicated subsidiary back in March 2025. The transmission gives a controller more modes to choose from. The AIHER system is the brain meant to choose between them, and that brain is still under development.
Range-extenders are having a moment, and FF wants the brain
Extended-range EVs are suddenly everywhere in the U.S. Scout, Ram and a string of other brands have all leaned toward the format. The pitch is simple: a battery big enough for daily driving, plus a small gasoline engine onboard to top up the pack on road trips. No charging anxiety, no highway range math.
What Faraday claims to add is decision-making. The premise is that an AI controller sitting on the decoupled hardware can pick the optimal mode — drive, charge, blended boost, full electric — faster than a rule-based hybrid controller, reading load, terrain, battery state and weather.
That last bit isn’t decoration. Faraday flags cold weather specifically, pointing to regions like the U.S. East Coast where lithium batteries shed usable range in deep cold. A range-extender that knows to lean on the engine earlier in a January cold-soak earns its keep.
The little engine doing the charging is its own rabbit hole — a Spanish shop has a crankshaft-free range-extender unit that’ll burn hydrogen or gasoline — but Faraday’s patent is about the gearbox logic, not the engine itself.
Faraday hasn’t exactly sold 1,000 cars yet
Here’s the reality check. By any honest measure Faraday is still a startup that has barely shipped vehicles. The FF 91 has dribbled out since 2023 in tiny numbers, the company shifted focus to the cheaper Super One under the FX brand it set up in late 2024, and the original Super One plan got reshuffled in May.
Per Faraday’s own May announcement, the first batch of mass-production Super Ones will be either an 800-volt battery-electric model or the AIHER hybrid, with the 400-volt project paused. The 800V route would put it in the same architectural club as the first 800-volt extended-range vehicles now coming out of China.
The production targets are where you raise an eyebrow. Faraday’s baseline business plan climbs from a few hundred cars to a quarter-million in five years.
Ambitious is one word for it.
The money tells its own story. Faraday raised $70 million across April and May, including a $45 million note agreement on April 17. But the company has been clear that the cash is earmarked for its robotics business, not car tooling — and the 800V/AIHER pivot is partly about freeing up resources for robots during their ramp.
Wall Street has noticed. FFAI trades deep in penny-stock territory, well under a dollar. Faraday picked up a Nasdaq minimum-bid deficiency notice in March, has until September 16 to claw back above $1.00, and stockholders approved a reverse split of up to 1-for-150 to dodge a delisting. Patents don’t fix any of that.
Series hybrid, parallel hybrid, and Faraday’s attempt at both
Strip the patent prose and the engineering reads like this. In a parallel hybrid — most Toyotas — the engine is mechanically wedded to the wheels through a planetary gearset, the motor chips in, and software optimizes at the margins.
In a series hybrid like the old BMW i3 REx or the upcoming Scout, the engine is fully divorced from the wheels and only ever spins a generator. Cleaner setup, but you eat conversion losses every time energy goes mechanical-to-electrical and back.
Faraday’s patent reaches for both. The engine can drive the wheels when that’s most efficient (highway cruising), generate when that’s smarter (city stop-and-go, cold-soak), or sit out when the battery has it covered. Decoupling hands the controller more options. Bolting AI on top is about making sense of those options dozens of times a second.
If the claim about simpler mechanics holds up, that’s a real manufacturing win. Hybrid transaxles are some of the most complicated objects in a modern car. Fewer hard linkages means fewer parts, less weight, and a shorter list of things that can fail at 80,000 miles.
A patent is not a product
A grant tells you Faraday owns the idea. It tells you nothing about whether the hardware survives a Michigan winter, costs less to build than a Toyota hybrid system, or behaves smoothly when an AI is flipping modes in heavy traffic. The design still has to be productionized, validated and dropped into a car someone will actually buy.
That car, if it shows up, is a Super One variant — and the timeline is now openly pinned to money. Faraday says deliveries start six to nine months after it secures major financing for the 800V model, nine to 12 months for the AIHER hybrid. Q2 2026 came and went without U.S. consumer deliveries.
And here’s the tell. Faraday spent June 2026 launching robots, not cars — a six-series lineup of humanoids and robot dogs, an education ecosystem, the works. The company now bills itself as a “Physical AI” outfit, and both the recent cash and the Super One pivot point toward the robotics side of the house.
YT Jia, Faraday’s founder and CEO, called the grant “central to FF’s vision of intelligent mobility.” Which is the kind of thing CEOs say about patents. The harder question is whether the AI-managed hybrid ever reaches a Super One a customer can drive home, and whether Faraday is still standing when it does. The engineering is interesting. The company shipping it is the variable.
