Three‑Year‑Old Robot Firm Zhiyuan (智元) Unveils Billion‑and‑Beyond Ambition, Sparks Shipment Dispute with Yushu (宇树)

Big reveal, bigger targets

A packed partner conference in Shanghai gave Zhiyuan (智元) — a robotics startup founded in 2023 — an unlikely stage to disclose its core operating metrics and grand ambitions. It has been reported that founder Deng Taihua, a former Huawei (华为) vice‑president who led Ascend AI business lines, said Zhiyuan’s revenue climbed from ¥300,000 in 2023 to ¥60 million in 2024 and then to ¥1.05 billion in 2025, and he publicly set targets for multiple‑fold growth in 2026, a ¥10+ billion goal in 2027 and a ¥100+ billion target by 2030. The tone was unmistakable: the company is trying to reframe the nascent "embodied intelligence" sector’s timeline from technology validation to mass deployment. But can such projections survive the scrutiny of auditors and markets?

Two rival playbooks collide

The public numbers have provoked an industry squabble with rival Yushu (宇树). Market research firm Omdia reportedly ranked Zhiyuan first in 2025 human‑form robot shipments with 5,168 units; Yushu countered, saying online figures were inaccurate and later disclosed in its prospectus that it shipped more than 5,500 human‑form robots in 2025 and turned a profit of ¥600 million that year. The dispute is not just about bragging rights. In an early market where benchmarks shape fundraising narratives and IPO valuations, “who’s number one” can materially affect partner confidence. Yushu’s path has been manufacturing‑led — self‑developed motors and reducers to crush bill‑of‑materials costs and achieve unit profitability — while Zhiyuan has pursued a platform and ecosystem strategy, pouring most R&D into AI and signing joint ventures and investments to build upstream and downstream ties.

Business model and credibility tests

It has been reported that Zhiyuan’s conference included a live eight‑hour demo where four robots completed 2,283 tasks with a claimed 100% success rate on a tablet assembly line — a proof point that robot capability is moving from “can it move?” to “can it do useful work?” Still, Zhiyuan acknowledged it remains loss‑making and must rely on scale to dilute R&D spend; its public roadmap reframes current losses as an entry cost for a later productivity payback. Investors and regulators will want audited financials and a prospectus that aligns marketing claims with accountancy — you can stage a partner gala, but a filing with the China Securities Regulatory Commission will be the truer test.

Strategic and geopolitical backdrop

All of this plays out against broader US‑China tech tensions and export controls on advanced chips and sensors that affect autonomous systems globally, factors that could constrain component access or raise costs for firms racing to scale. Zhiyuan’s Huawei‑style ecosystem playbook arguably aims to lock supply and demand together; Yushu’s cost engineering is a simpler, more defensible route if component shortages bite. In short: China’s embodied‑AI market is still a wild frontier, and today’s headline numbers will determine who leads the narrative — until auditors and markets decide who leads the market.