Embodied Intelligence Observer
墨奇智能 Logo

Morphi墨奇智能

LatestAngel series (over RMB 1B, jointly led by Alibaba and Tencent)

Post-money over RMB 7B, a China embodied intelligence first-round record among publicly disclosed deals

Shenzhen Morphi Intelligence Technology Co., Ltd.

Founded by Huawei's intelligent-driving brain trust; a software-hardware integrated general humanoid robotics company; Alibaba and Tencent jointly led a rare co-investment; first product released in July 2026 for commercial service scenarios

Humanoid RobotsChinaWebsite

数据截至2026-08-19

FoundedAug 2025 (registered in the Qianhai Shenzhen-Hong Kong Cooperation Zone, Shenzhen)
HQShenzhen Qianhai
FounderGao Wenli, Huang Qingqiu
Products5 items
Funding Rounds1 rounds
Events11 items

About

Morphi was registered in Shenzhen Qianhai in August 2025. Founder/CEO Gao Wenli is an 11-year Huawei overseas veteran and co-founder of cross-border logistics company iMile; co-founder/CTO Huang Qingqiu is a Huawei 'Genius Youth' who headed the smart-driving AI department of Huawei's Intelligent Automotive Solution BU, steering ADS 1.0-to-4.0 full-chain algorithm iteration and million-unit production — the first in the industry to bring one-stage end-to-end (the WEWA architecture) to million-scale production. The company positions itself as a software-hardware integrated general humanoid robotics company, migrating autonomous driving's data loop, on-device model optimization and production quality control into robotics; embodied manipulation lead Lin Tianwei is a core proposer of the UniAD end-to-end driving paradigm. The first product was released in July 2026 for commercial service scenarios (confirmed not to enter homes directly), starting with hotels and serviced apartments, with hotel-scenario POCs starting in H2 2026. In July 2026 it announced over RMB 1B in angel-series funding jointly led by Alibaba and Tencent — a rare pairing — at a post-money valuation over RMB 7B, setting a China embodied intelligence first-round record among publicly disclosed deals.

Technical Approach

Systems-engineering route: migrates the data-loop logic validated on millions of vehicles, on-device model optimization, multi-sensor fusion and production quality control from intelligent driving into robotics. A methodology of 'data quality first, model architecture second'; the data flywheel is wearable collection → AutoLabel auto-annotation → training → multi-line-of-defense scaled evaluation → feedback into collection. Three-stage model path: post-training (done) → foundation-model training on open-source models → training a native embodied model from scratch within 2026 (starting at 3-7B parameters); it deliberately avoids the 'world model' or 'VLA' labels. It rejects the 'industrial data feeding home intelligence' path, focusing on relatively hard commercial scenarios like hotels and apartments.

Commercialization(5)

  • First product released in July 2026 (a service-scenario robot; full form-factor specs not all disclosed)
  • Hotels and serviced apartments first: self-developed wearable collection devices already collecting data with hired cleaners in real scenarios
  • Hotel-scenario POC starting in H2 2026; after validating the single-store model, push scaled rollout
  • Continuously finding vertical scenarios to land capabilities accumulated in hotel scenarios
  • As of the funding disclosure, no customers, orders, deliveries or revenue had been publicly disclosed

Position & Strengths(5)

  • Top team: Huawei 'Genius Youth' Huang Qingqiu (ADS 1.0-4.0, million-unit WEWA) + Huawei veteran and unicorn co-founder Gao Wenli + UniAD proposer Lin Tianwei — all with rare 'built systems from zero to mass production' track records
  • Capital endorsement: Alibaba + Tencent jointly leading an over-RMB 1B angel round at an over-RMB 7B valuation, a China embodied first-round record
  • Data methodology: the AD data flywheel (collect - auto-label - evaluate - iterate) migrated wholesale, with a data-quality-first philosophy ahead of the industry
  • Software-hardware integration: robot body, embodied foundation model and cross-scenario generalization engine all in-house
  • Pragmatic scenario choice: commercial first, home later — prove commercialization before going consumer

Links1

Morphi | Embodied Intelligence Observer