Company
Robot / product
G1, H1, H2 and R1 humanoids; Go2, B2 and A2 quadrupedsPublic
Funding
Series C, June 2025: ~¥700M at a ¥12B+ post-money (co-led by China Mobile fund, Tencent, Jinqiu, Alibaba, Ant Group, Geely Capital). STAR Market IPO Aug 19, 2026: issue price ¥150.8, raised ¥6.1B against a ¥4.2B target; shares opened at ¥1,100Public
Target market
Research labs, developers, education, entertainment; industrial inspectionPublic
GTM
Target vertical
Universities and labs, developers, media and entertainment, inspectionPublic
ICP
Anyone who wants capable hardware at a fraction of Western pricesAnalyst view
Primary buyer
Lab directors, developers, distributors, consumersAnalyst view
Use case
Research platform, demos and performances, patrol and inspectionPublic
Sales motion
Low-price direct and online sales (R1 on AliExpress Brand+ in North America, Europe, Japan, Singapore from April 2026) plus dealers such as RoboStore in the USPublic
Deployment model
Ship the product; customer integratesPublic
Commercial model
CapEx product salesPublic
Pricing
G1 launched at $16,000 (May 2024), since listed from $13,500; Go2 from $1,600; R1 from $5,900 (July 2025; ¥29,900 in China). US dealer RoboStore: R1 Basic $8,990, EDU from $15,950, EDU Pro from $28,900Public
Sales cycle
TransactionalPublic
Intelligence
8/10MARKET FIT
8/10DISTRIBUTION STRENGTH
8/10BUYER READINESS
Commercialization riskMedium: FCC Covered List limits the US to pre-authorized models; US share shrinking; US trust sits with dealers
Key GTM insightUnitree sells hardware the way consumer electronics are sold: published price, online checkout, distributors. It owns the research and developer market because it is the cheapest way to get a capable body.
What competitors should learnThe developer market is a distribution channel. Every lab running your robot trains the engineers your enterprise customers will hire.
Provenance: Public = press, company site or filings · Analyst view = ANSIOLABS inference from public evidence · Not disclosed = we do not know and will not guess. Scores are analyst assessments, methodology v0.1