This article assembles the verified quantitative data on China’s humanoid robot industry in 2025-2026, funding totals, unit deliveries, revenue figures, production targets, and technology benchmarks, to give a calibrated picture of where the industry actually is versus where the headlines suggest it is. The humanoid robot boom is real. The scale in 2025 is also real but modest, tens of thousands of units, hundreds of millions of dollars in revenue, and early-stage AI that works in structured environments. The 2028-2030 projection is where the numbers become genuinely large. Understanding the gap between current reality and projected scale matters for anyone evaluating humanoid robot investments or deployment decisions.
Units Delivered: 2025 Reality
Verified 2025 humanoid robot deliveries by Chinese manufacturers (based on company disclosures and industry reports): AgiBot 10,000 units by March 2026 (most commercially advanced), Unitree G1 several thousand units (IPO filing implies 2,000-5,000 units in 2025), UBTECH Walker S2 500+ units (mass production November 2025, first batch target), Fourier GR series 150-500 units, Engine AI SE01/PM01 150-500 units, Galbot G1 few hundred units, Leju KUAVO 577 units (disclosed). Total verified Chinese humanoid deliveries in 2025: approximately 12,000-18,000 units. MIIT target for 2025: 10,000 units, exceeded by AgiBot alone.
For context: Chinese industrial robot deliveries in 2025 were approximately 490,000 units. Humanoid robots at 12,000-18,000 units are approximately 3% of total robot deliveries in China, significant for a new product category in its first commercial year, negligible as a fraction of total robotics production. The humanoid market is currently an order of magnitude smaller than the news coverage implies relative to industrial robotics.
Funding: The Billion-Dollar Race
Disclosed funding as of mid-2026: AgiBot $1B+ (multiple rounds including $300M Series B, December 2024). Galbot $300M (December 2025, single round). Leju KUAVO CNY 1.5B ($207M, October 2025). Unitree undisclosed (IPO filing will disclose). UBTECH listed on HK (total raised $820M across all rounds including pre-IPO). Fourier $100M+ (Series C and later rounds). Engine AI $100M+ estimated. Zhiyuan CNY 1B ($138M) estimated. Total disclosed Chinese humanoid funding in 2024-2026: approximately $3-4 billion. Total disclosed plus estimated: $5-7 billion. This is concentrated in 6-8 companies. The median Chinese humanoid startup has raised $20-50M and is not in the public discussion.
Production Targets: Government vs Market
MIIT targets: 10,000 units in 2025 (achieved), 50,000 in 2026, 200,000 in 2028. These are aspirational government KPIs, not contracted orders. Market-based projections from IDC, Goldman Sachs, and Morgan Stanley: 50,000-100,000 units per year globally by 2028, with China producing 60-70% of units. AgiBot 10,000 units by March 2026 alone suggests 2026 could reach 50,000+ units from AgiBot’s production program. Efort super plant phases: 50,000 industrial robot units per year by 2027 (not humanoids, but indicates total Chinese robot manufacturing capacity growth). The MIIT 2026 target of 50,000 humanoid units is achievable if AgiBot reaches its stated commercial trajectory.
Technology Benchmarks: What Works Today
Manipulation: AgiBot G2 claims sub-millimeter assembly precision in structured environments. Unitree UnifoLM-VLA-0 achieves 91.5% task success on VIRAL benchmark. UBTECH Walker S2 handles 5 kg per hand with 0.5 N force resolution. Current AI: works reliably in controlled, previously-characterized environments. Generalization to novel objects and environments is the capability gap. Commercial deployment contexts where humanoids work: battery factory material logistics (AgiBot-CATL), warehouse picking (structured, repetitive), machine tending (fixed workstation), quality sampling (known objects). Contexts where humanoids are not yet deployed commercially: unstructured environments, outdoor logistics, full assembly line automation.
What to Watch in 2026-2027
The inflection points that will determine if the projected scale is achieved: (1) AI generalization, can humanoid AI handle novel objects and environments without re-programming? (2) Reliability at production duty cycle, 2,000+ hours per year operation without failure. (3) Safety certification, MIIT and SAC standards enabling certified deployment in regulated environments (medical, food, aerospace). (4) Price erosion, will humanoid prices fall below CNY 200,000 ($27,500) for commercial deployment? AgiBot has already moved from $170K to $95K for A2 in volume. The trajectory suggests CNY 200,000 is achievable by 2028.
Complete company profiles: Top 40 China Robot Rankings 2025. For the humanoid leaders: AgiBot, Unitree, UBTECH.
Sources
- IDC: China humanoid robot market report 2025
- MIIT: Robot industry development plan and KPI tracker 2025
- Goldman Sachs: Humanoid robotics market projection 2024
- Company disclosures: AgiBot, Unitree, UBTECH, Galbot, Leju, Fourier (2025-2026)
