EngineAI

EngineAI is a Shenzhen humanoid-robotics startup founded in 2023, distinguished in the Chinese cluster by gait-and-motion-control engineering on the SE01 and PM01 platforms.
EngineAI

EngineAI, trading as 引擎人工智能 (Yinqing Rengong Zhineng), is a Chinese humanoid-robotics company headquartered in Shenzhen, founded in 2023 by Zhao Tongyang (赵同阳), an engineer with previous experience at multiple Chinese robotics and automation companies. EngineAI develops the SE01 and PM01 humanoid platforms, with a strategic emphasis on gait-and-motion-control engineering that has produced some of the most-watched humanoid running and dynamic-locomotion demonstration videos in the Chinese humanoid cluster as of 2025.

At a glance

  • Founded: 2023 in Shenzhen by Zhao Tongyang.
  • Status: Private. Multiple funding rounds completed; cumulative disclosed funding not itemised.
  • Funding: Multiple private rounds from Chinese venture firms. Specific investor list less publicly itemised than at the larger competitors.
  • CEO: Zhao Tongyang, founder.
  • Open weights: None publicly released for the platform AI stack. Hardware specifications have been published in product-launch materials.
  • Flagship products: SE01 (full-size humanoid platform, 2024); PM01 (smaller humanoid form factor, 2024 to 2025); related demonstration-platform prototypes.

Origins

Zhao Tongyang founded EngineAI in Shenzhen in 2023 with an engineering team drawn from Chinese industrial-robotics and consumer-electronics manufacturing backgrounds. Shenzhen's mechanical-and-electronics supply chain, the densest in China for the kind of components humanoid-robotics platforms depend on, was a natural choice of headquarters for a hardware-focused humanoid startup at the entry point of the Chinese cluster's expansion.

The company's strategic focus from founding has been on gait-and-motion-control engineering, with the public demonstration cadence built around progressively more-difficult dynamic-locomotion videos: walking with payload, running at human-like speeds, recovery from disturbances, and stair-and-rough-terrain navigation. The demonstration videos have circulated heavily on Chinese social platforms (Weibo, Douyin, Bilibili) and on international platforms (Twitter/X, YouTube), with multiple individual videos drawing tens of millions of views each.

The SE01 platform was the company's first full-size humanoid demonstration, introduced in 2024. The PM01 followed in 2024 to 2025 as a smaller form-factor variant, with the explicit positioning as a lower-cost research-and-developer SKU. The company's manufacturing capacity in Shenzhen has expanded through 2024 and 2025, with reported initial production runs of several hundred units across the SE01 and PM01 lines.

The competitive distinctiveness of EngineAI's positioning has been the gait-and-locomotion videos, which have been notably more sophisticated than the early demonstration material from most other Chinese cluster entrants. The technical capability behind those videos has been a frequent subject of speculation in Chinese technology media, with the consensus framing crediting the company's engineering team rather than any single distinctive technology breakthrough.

Mission and strategy

EngineAI's strategic positioning is on dynamic-locomotion engineering and on the demonstration-driven brand-building that follows from compelling motion-control video content. The company has been notably effective at the social-media-and-technology-press demonstration cycle, with each new demonstration video generating coverage in international technology media in ways that the larger but less video-savvy competitors have not consistently matched.

The product strategy has been to develop two humanoid form factors: the SE01 full-size platform and the PM01 smaller-form-factor platform. The dual-SKU approach gives the company coverage of both the academic-research market (where smaller-and-cheaper platforms are preferred) and the industrial-pilot market (where full-size platforms are required). The pricing has been less publicly itemised than at Unitree, but the positioning has been clearly in the lower-cost tier of the global humanoid market.

The AI-stack integration has been less publicly visible than at Agibot or the US cohort. The company's public materials have emphasised the mechanical-and-control-engineering capability rather than the foundation-model integration, suggesting that the AI stack is either still in development or is being sourced through partnership rather than developed in-house at the same intensity. The watchable evolution over the next 12 to 18 months is whether EngineAI develops its own AI-foundation-model capability, partners with one of the Chinese frontier labs (DeepSeek, Alibaba Qwen, Moonshot AI), or positions itself as a hardware platform that other AI providers integrate with.

Models and products

  • SE01 (2024). Full-size humanoid platform. Approximately 1.7 metres tall, with the company's proprietary dynamic-gait control stack. Demonstration videos have shown running, stair-navigation, and disturbance-recovery capability at a level competitive with the leading global research-humanoid platforms.
  • PM01 (2024 to 2025). Smaller form-factor humanoid platform. Lower-cost research-and-developer SKU targeting academic and individual-research customers.
  • Demonstration prototypes. Additional internal-development prototypes have appeared in company demonstration material but have not been publicly named or priced. The cadence suggests an active development pipeline beyond the SE01 and PM01 product line.

Benchmarks and standing

EngineAI's standing in the Chinese cluster is measured primarily through the demonstration-video cadence and through the production-and-deployment volume of the SE01 and PM01 platforms. The demonstration videos have been distinctive in the Chinese cluster for their gait-and-locomotion sophistication, and the company has consistently been cited in Chinese technology media as among the more-impressive Chinese humanoid entrants on the motion-control dimension specifically.

Production volumes have not been officially itemised. Industry coverage in 2025 placed cumulative SE01 and PM01 shipments in the low-to-mid hundreds of units, primarily into research-and-development customer pipelines. The competitive position relative to the cluster leaders (Unitree, UBTECH, Agibot) is clearly behind on volume, but the company's brand recognition in the technology-press space has been disproportionate to the production volume.

The AI-foundation-model capability has been the less-visible side of the company. The motion-control demonstrations have been impressive, but the broader manipulation, task-completion, and general-purpose AI capability that would translate to industrial deployment has been less publicly evidenced. Whether this is a function of the company's communications strategy (emphasising the most-impressive capabilities while keeping the under-development capabilities private) or a function of the underlying capability gap is the central watchable question.

Leadership

  • Zhao Tongyang (Founder, Chief Executive Officer). Engineering background with previous experience at multiple Chinese industrial-robotics and consumer-electronics companies. Public profile has been lower than the founder profiles of Peng Zhihui at Agibot or Wang Xingxing at Unitree, but the company's demonstration cadence and engineering output have positioned EngineAI alongside those companies in the Chinese cluster framing.
  • Senior engineering team. Approximately 200 to 400 employees as of late 2025, with the senior engineering organisation concentrated in mechanical, actuator, and control-systems engineering. The company has been less publicly visible on AI-foundation-model and software-engineering hiring than the AI-emphasising competitors.

Funding and backers

EngineAI has raised multiple private rounds from Chinese venture firms. The specific named-investor list has been less publicly itemised than at the larger Chinese humanoid competitors. Industry coverage has placed the company in the mid-tier of the Chinese cluster on cumulative funding, behind Unitree, UBTECH, and Agibot but ahead of multiple smaller competitors.

The company has not filed for IPO as of the current date. The funding trajectory will be informative for the company's competitive standing, particularly any named-investor disclosures that include strategic backers from the broader Chinese technology ecosystem (Tencent, Alibaba, ByteDance, Xiaomi, BYD have all participated in humanoid-cluster funding rounds for competitor companies).

Industry position

EngineAI is a tier-two competitor in the Chinese humanoid cluster, positioned on the demonstration-and-motion-control axis with a strong technology-press brand recognition relative to its production volume. The competitive position behind the three primary cluster leaders (Unitree, UBTECH, Agibot) is consistent across industry framings. The upward-repositioning path depends on (1) whether the demonstration capability translates to production-volume scaling, (2) whether the AI-foundation-model capability catches up to the motion-control capability, and (3) whether the company secures named-strategic-customer or named-large-investor partnerships that materially shift the funding-and-credibility position.

The competitive question against Unitree is around cost positioning at volume, where Unitree's manufacturing scale gives it a structural advantage. The competitive question against UBTECH is around enterprise-customer pipeline, where EngineAI has been less visible. The competitive question against Agibot is around AI-foundation-model integration. The competitive question against Fourier Intelligence is around strategic-partner positioning, where Fourier's Saudi relationship gives it a distinctive asset.

Internationally EngineAI competes in the research-and-developer humanoid market against the lower-cost SKUs of the US cohort and against the broader Chinese cluster on price-and-capability positioning. The company's competitive position in international export markets has been less publicly mapped than for the larger Chinese cluster competitors.

Competitive landscape

Outlook

Open questions and watchable signals over the next 6 to 18 months:

  • Production-volume scaling. SE01 and PM01 shipments through 2025 have been in the low-to-mid hundreds. Scaling to thousands annually will be the key indicator of whether the company can convert demonstration capability into market-relevant volume.
  • AI-foundation-model capability disclosures. The motion-control capability has been the company's headline differentiator. Whether the AI stack catches up to the mechanical capability, and whether the company partners with a Chinese frontier lab or develops the AI capability in-house, will be the central watchable strategic question.
  • Named-customer-and-investor pipeline. EngineAI has been less visible than competitors on named-customer announcements (UBTECH's automotive partnerships, Agibot's enterprise pilots) and on named-large-investor rounds. Movement on either dimension would materially shift the competitive position.
  • Next-generation product launches. SE01 and PM01 established the dual-form-factor product strategy. Successor generation announcements in 2026 will indicate whether the company maintains the product-development cadence required to compete in the cluster.
  • Funding round timing and disclosures. Industry coverage has placed EngineAI in the mid-tier of the cluster on funding. A named large round or strategic-partner announcement in 2026 would be a meaningful upward-positioning signal.

Sources

About the author
Nextomoro

Nextomoro

nextomoro tracks progress for AI research labs, models, and what's next.

AI Research Lab Intelligence

nextomoro tracks progress for AI research labs, models, and what's next.

AI Research Lab Intelligence

Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to AI Research Lab Intelligence.

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.