After 12 years and roughly $3.2 billion in cumulative funding (per Tracxn data), Faraday Future has delivered fewer than 20 vehicles. Now the Jia Yueting-led company is abandoning the EV label altogether: an August shareholder meeting will vote on a renaming that rebrands FF as a “physical AI ecosystem company,” with humanoid and embodied-AI robots as the core business. A Chinese industry analysis by OFweek argues the financial model behind the pivot doesn’t hold up.
The robot numbers first. FF says it delivered 242 robots by end of June — 22 in March, 46 in April, 69 in May, 105 in June — and has raised its full-year shipment target to 2,000 units. Hitting that target requires averaging 300+ units a month for the rest of the year, triple its best month to date.
The business model is the sharper issue. FF does not design or manufacture these robots. It purchases finished units from Chinese makers, rebadges them, applies light software adaptation, and resells them to US B2B customers — schools, government and enterprise buyers — earning a hardware markup plus basic software service fees. OFweek’s assessment is blunt: the model is functionally identical to a cross-border distributor, with three structural ceilings — full supply-chain dependence on Chinese OEMs, no technology moat, and end customers who can eventually buy direct from the original manufacturers.
FF touts its robot line as achieving positive per-unit gross margin — the company’s first profitable product in 12 years. But per-unit gross margin covers only direct procurement and adaptation costs. FF posted an operating loss of $185 million in 2025, with annual losses running near $400 million. Its $70 million in dedicated robot funding (including a $45 million April raise from a US institutional investor) covers roughly half a year of operating burn.
Industry breakeven math cited in the analysis: FF would need to sell 5,000–10,000 robots a year to offset its losses. Even at the full 2,000-unit target, annual robot revenue of about $40 million would cover less than a third of the company’s burn. The software-premium thesis also struggles: leading Chinese makers like Unitree and AgiBot ship with full self-developed software stacks, vision models and OTA services, leaving a rebadger little room to add value. FF’s software revenue is about 26% of the robot line — mostly one-off licence sales rather than recurring subscriptions.
The backdrop, as reported in the piece: Jia remains subject to enforcement actions in China totalling RMB 2.27 billion in personally executed amounts, with an additional RMB 2.61 billion in resumed enforcement added in June, and an active consumption-restriction order. His stated plan — build the AI ecosystem, repay debts, return to China — now rests entirely on a distribution business whose economics, on OFweek’s arithmetic, cannot get there.
For Western investors, the FF story is a useful lens on a bigger asymmetry: Chinese humanoid robots are now good enough, and cheap enough, that reselling them in the US is a viable entry-level business — just not a venture-scale one.
Read the original report (OFweek Robotics)
*Translated and adapted from OFweek Robotics (https://robot.ofweek.com/2026-07/ART-8321203-12003-30694818.html).*