The image of human‑shaped robots dancing on China’s Spring Festival Gala has driven a sudden, visible surge in demand for humanoid machines — not for purchase, but for hire. Since early 2026 enquiries to rental firms have multiplied, and market researchers now estimate a near tenfold increase in China’s robot‑rental sector from roughly ¥1bn in 2025 to as much as ¥10bn–¥100bn this year, depending on methodology. That spike has put a spotlight on a fragile but fast‑forming commercial pathway by which embodied artificial intelligence meets mass audiences.
For many clients — event planners, shopping malls, corporate galas and a handful of novelty‑hungry parents — renting a robot makes more sense than buying one. Retail prices for flagship humanoids remain in the tens of thousands of dollars, technology cycles are rapid, and early buyers risk quick obsolescence. Renting bundles the device with choreography, transport and on‑site technicians: today’s market is less about standalone machines than “robots plus service.”
The economics are changing quickly. Daily rental fees that averaged ¥10,000–¥20,000 in early 2025 have fallen to ¥3,000–¥5,000 for many bookings in 2026, with some platforms advertising sub‑¥1,000 experience packages. Supply‑side expansion — driven by 2025’s mass production push and the launch of marketplace platforms — is pushing prices down even as public exposure raises demand. Platforms such as the newly founded “Qingtianzu” are standardising delivery, pricing and fault handling, aiming to turn a dispersed, bespoke trade into repeatable transactions.
Operational realities, however, expose the sector’s current immaturity. Most performances still depend on remote human operators: one technician per machine is common for synchronised dances, and robots typically work just two to three hours on a charge. Rental firms therefore amortise costs by pairing equipment with technical crews and spare units for rotation, creating a labour‑intensive, capital‑heavy business model. In practice the offering is Robot‑as‑a‑Service rather than a substitute for human performers.
A more acute risk appears on maintenance and warranty timelines. Many first‑wave machines shipped in 2025 come with one‑to‑18‑month warranties. From 2026 a substantial portion of fleets will step beyond warranty, exposing rental operators to high repair bills, steep depreciation — industry figures cite annual write‑downs of 20–30% — and unpredictable downtime. For some rental businesses, a single major repair can erase the margin of dozens of bookings.
These pressures have shaped company strategy: some robot makers focus on industrial and logistics uses where higher per‑unit margins and clearer productivity gains justify heavier R&D investment, while others treat rental and entertainment as cashflow routes that boost brand recognition. The middle line of argument among analysts is that renting is simultaneously a transitional diffusion mechanism and a potentially enduring commercial niche: it accelerates public familiarity while buying time for autonomy and robustness to catch up.
For investors and policymakers the sector offers both promise and warning. The rapid consumer‑facing exposure of humanoid robots creates brand and market momentum, but long‑term viability depends on engineering advances — longer battery life, stronger autonomy that reduces operator headcount, and lower maintenance costs — together with standardised after‑sales frameworks. How the industry navigates the warranty cliff and the cost squeeze this year will determine whether robot rental becomes a durable business model or a transient spectacle that fades as novelty diminishes.