美西时间10月12日，FF EAI Robotics和Faraday Future把最终合并协议签了。FFR用全股票方式收下FFAI的机器人资产和业务，发行价每股2.246美元，机器人业务整体估值约2亿美元，相对签约日前收盘价有约50%到55%的溢价。交易过了双方利益不相关独立董事组成的特别委员会，独立财务顾问出具公允性意见，目标本月底前走完交割。
On October 12 in US Pacific time, FF EAI Robotics and Faraday Future signed the definitive merger agreement. FFR acquires FFAI's robotics assets and business in an all-stock deal at 2.246 per share, valuing the robotics business at about 200 million, a roughly 50% to 55% premium to the prior closing price. The deal passed special committees of independent directors on both sides, with a fairness opinion from independent financial advisers, and targets closing by the end of October.
这笔交易更像把机器人业务单独装进一个上市容器，而不是普通并购补产品线。FFR交割前按充分稀释口径股权价值约6000万美元，同步签了不超过1000万美元首轮PIPE，按市价发行，没折扣也没认股权证。交割后FFAI拿约65.7%，新增股份锁18个月；管理层和员工约18.2%；PIPE投资人约5.5%，六个月或转售注册生效前不能卖；原FFR其他股东约10.7%。扣掉受限部分，自由流通公众股大约1300万股，占10.7%，盘子薄得经不起大资金蹭一下就晃。
This looks more like putting the robotics business into a standalone public vessel than a normal acquisition topping up product lines. Before closing, FFR's fully diluted equity value is about 60 million, with a first PIPE of up to 10 million at market price, no discount and no warrants. After closing, FFAI holds about 65.7%, new shares locked 18 months; management and employees about 18.2%; PIPE investors about 5.5%, restricted six months or until resale registration takes effect; other legacy shareholders about 10.7%. After stripping restricted stock, free float is around 13 million shares, 10.7%, thin enough that any headline moves the quote.
FFR前身是AIxC，从技术基础设施、AI agent一路切到EAI机器人生态，这次把代码和定位一起换掉。FFAI把机器人资产挪出去，自己往Robotaxi共享网络、AI视舱运营商、Physical AI投资控股公司靠。汽车业务不再只围着车型交付转，往后三条线分别是运营Robotaxi并接入外部网络、把智舱系统外拓到其他智能车、让FF自有车型进网络。机器人单独融资，整车保留控股入口，账先分开算，故事才能分别讲。
FFR's predecessor was AIxC, moving from technical infrastructure and AI agents into the EAI robotics ecosystem, and this deal swaps both the ticker and the positioning. FFAI moves the robotics assets out and repositions toward a Robotaxi sharing network, AI cockpit operator and Physical AI investment holding company. The car business stops revolving only around vehicle delivery. The three lines are operating Robotaxi and connecting to external networks, pushing the cockpit system to other smart vehicles, and putting FF's own models into the network. Robotics finances separately, the auto side keeps control, the books split first and the narratives split with them.
眼下最硬的数据还是交付。FFAI侧已经铺出三大形态、五大系列、二十多款本体产品，截至2026年9月底EAI机器人本体累计销售及出货817台，公司口径强调正毛利交付。 817台放进全球机器人产业里不算大批量，但足够把「只有样机」和「真往外发货」切开。看这个数别只看台数，要盯签收单、回款周期、售后工单和现场停机时间，机器人出了厂门才刚开始消耗现金流。
The hardest data right now is still delivery. FFAI has laid out more than 20 body products across three forms and five series. By the end of September 2026, cumulative EAI robot sales and shipments reached 817 units, with the company emphasizing positive gross-margin delivery. Eight hundred and seventeen units is not mass volume globally, but it separates having only prototypes from actually shipping machines. Do not read the unit count alone. Watch signed receipts, collection cycles, service tickets and on-site downtime. A robot starts burning cash after it leaves the door.
五年预测写得满，但层级要拆开。公司管理层初步预测2026到2030年累计收入约19.8亿美元，机器人本体五年累计销量超13万台，累计数据供给超1900万小时，非本体生态收入占比从2026年22%抬到2030年接近50%。 这类数还没被当成已采纳业绩指引，别直接折进估值模型。硬件卖一台是一台的毛利，数据、租赁、运维、教育、Roboshare协同才是后期能不能撑起2亿美元估值的另一半。前者养活团队，后者才让美股愿意把它当Physical AI平台而不是设备经销商。
The five-year numbers are full but need layers. Management's preliminary forecast puts cumulative revenue at about 1.98 billion from 2026 to 2030, cumulative robot body sales above 130,000 units, cumulative data supply over 19 million hours, and non-body ecosystem revenue rising from 22% in 2026 to nearly 50% in 2030. These are not yet adopted guidance, so do not drop them straight into a valuation model. Hardware sale is per-unit gross margin; data, leasing, operations, education and Roboshare synergies are the other half that justifies a 200 million label. Hardware feeds the team, the ecosystem is what makes US markets treat it as a Physical AI platform instead of an equipment dealer.
轻负债是这次重组故意留出来的状态。交割后FFR账面几乎不背债，经营目标最晚2028年三季度经营性现金流转正，此前累计融资需求不超1.5亿美元。这个窗口期很短，机器人公司最容易死在样机漂亮、量产卡壳、售后铺不开的阶段。PIPE只开1000万美元额度，用意不是把钱一次拿满，而是先把上市主体撑住，避免过早稀释，后面看交付节奏再补弹夹。
Near-zero debt is a deliberately arranged state for this restructuring. After closing, FFR carries almost no liabilities, with operating cash flow targeted to turn positive no later than Q3 2028 and cumulative funding needs under 150 million before that. The window is short. Robotics companies most often die when prototypes look good but volume production stalls and service coverage lags. Keeping the PIPE at up to 10 million is not about maxing proceeds, it holds the listed entity together, limits early dilution, and leaves room to reload once delivery cadence is visible.
流通盘薄会把消息面放大成股价波动。1300万股自由流通对应约10.7%，短线资金对交割公告、PIPE到账、新品发布会比季度毛利更敏感。FFAI控股65.7%并锁18个月，等于先把筹码压住，给业务跑数据留时间。但这也意味着盘面成交价不代表整家公司真实定价，一小撮流通股在替2亿美元估值报价。锁定期、转售注册、后续融资批次要比K线更值得盯。
A thin float turns headlines into price swings. About 13 million free shares, 10.7%, makes short-term money more sensitive to closing notices, PIPE funding and product launches than to quarterly gross margin. FFAI's 65.7% stake locked for 18 months presses the chips down and buys time for operating data. But traded price is not real pricing for the whole company. A small float is quoting the $200 million valuation. Lock-ups, resale registration and later financing batches matter more than the candle chart.
FFAI剩下的主线往投资控股和出行运营收。Robotaxi、AI视舱、Cybercab接入这些词听着远，落到财务上就是车队运营收入、授权收入、调度和网络服务费。机器人本体拆出去独立上市，等于把高波动、高叙事、重交付的业务放进容错率更高的壳里，整车侧不被机器人研发节奏拖着补血。两边要靠场景打通，Roboshare把舱内、车端、机器人端的数据和任务接起来，不然还是各讲各的PPT。
FFAI's remaining thread pulls back to investment holding and mobility operations. Robotaxi, AI cockpit and Cybercab access sound far out, but on the financials they mean fleet operating revenue, licensing revenue, dispatch and network service fees. Spinning robotics out to a separate listing puts the high-volatility, narrative-heavy, delivery-heavy business into a shell with more tolerance, so the vehicle side is not dragged into repeated fundraising by robotics R&D节奏. The two sides have to share scenarios. Roboshare links cabin, vehicle and robot-side data and tasks. Without that, they remain separate decks.
产品发布紧跟着交易节奏走。FF Nova、FF Nova Mini和开发者强化学习平台把第一波用户指向教育机构、开发者和实验室，这条路比直接卖工厂产线稳，单机故障影响小，课程和Skills产品能把留存做厚。教育市场未必撑得起13万台总量，但能先把交付链路、固件更新、遥控接管、数据回收跑顺，再往商业服务和高动态场景推，比一上来喊人形进家庭靠谱。
Product launches ride right behind the transaction. FF Nova, FF Nova Mini and the developer reinforcement-learning platform point the first users at schools, developers and labs. This path is steadier than selling straight into factories. A single unit failure has smaller blast radius, and courses plus the Skills product thicken retention. Education alone will not carry 130,000 units, but it lets delivery, firmware updates, teleoperation, takeover and data recovery run clean before pushing into commercial service and high-dynamic scenes. Promising humanoids in homes from day one is cheaper talk.
核验顺序别被发布会带跑。先等10月底前交割文件和8-K落地，再看PIPE资金实际到账，接着对账机器人流向，客户签收、租赁归还、退机都要拆出来，最后才看Physical AI、Robotaxi这些大词有没有对应收入科目。817台是起点，不是护城河。EAI这行先把售后和数据链路跑正，才有资格谈平台化，否则2亿美元估值只是上市壳提前把未来五年贴现了。
Do not let the launch distract the check sequence. Wait for the closing documents and 8-K by end of October, then confirm PIPE funds actually land, then reconcile robot flows, separating customer sign-off, lease returns and returns. Last, see whether big words like Physical AI and Robotaxi map to revenue lines. 817 units is a start, not a moat. In EAI you only earn the right to talk platforms after after-sales and data links run straight, otherwise the $200 million valuation is just a listed shell discounting the next five years early.
