What This Is

Last week, NetEase Yunxin partnered with Lingxi to embed an AI brain into two-wheeled EVs—93% outdoor wake-up rate, mesh intercom (vehicles auto-link into a small network without carrier base stations) up to 1.5 km, and functional offline. What's worth paying attention to is that the next battleground for AI-ified hardware has shifted from the phone screen to the wheels.

The solution, called "Zhiyu Suixing" (Smart Voice On-the-Go), runs on Lingxi's noise-cancellation, main control, and mesh networking chips on the hardware side, paired with NetEase Yunxin's RTC (real-time audio/video) backbone and its proprietary AI hardware agent platform on the software side. It has already been integrated with leading OEMs including Xinri, Niu, Ninebot, Yadea, Shouqu, and shared-mobility operators.

Compliance is another selling point: it embeds NetEase Yidun content security, covers China's "July 15 Regulation" (Interim Measures for the Management of Anthropomorphic AI Interactive Services), the EU GDPR, and the EU AI Act, among other regulatory regimes. Overseas, it deploys independent multi-center data nodes with sensitive data stored locally and never crossing borders.

Industry View

Bulls see this as a landmark sample of AI landing in non-phone hardware. Two-wheeled EVs are shifting from mere transportation to "emotional carriers": from 2025 to 2026, cycling-related short-video views on Xiaohongshu exceeded 8.2 billion; 47% of core users aged 18–30 have joined night rides or group rides; 71.5% of young buyers prioritize smart features when purchasing. For OEMs, the solution compresses in-house compliance cycles from 6–12 months to roughly 1 month—a substantial efficiency win.

But we note several cautionary counterpoints:

  • This is a B2B solution, not a product consumers can buy directly; the end-user experience depends on whether OEMs actually promote it as a differentiating feature.
  • Leading brands like Xinri, Niu, and Ninebot already have their own in-vehicle systems; "Zhiyu Suixing" is more a supplement than a replacement, leaving OEMs with limited leverage.
  • Two-wheeled EVs are priced mainly in the RMB 2,000–6,000 range. Whether the BOM and traffic costs of edge chips plus cloud LLMs can be amortized in low-end tiers will determine if the solution can truly scale.
  • Always-on in-vehicle voice monitoring creates new privacy boundary issues; passing compliance review does not mean genuine user acceptance.

Impact on Regular People

For Enterprise IT: OEMs and shared-mobility operators gain a packaged AI + compliance + global connectivity solution that significantly cuts in-house R&D investment, but in return they cede some data sovereignty.

For Individual Careers: Roles with hybrid "AI plus hardware" backgrounds will become more valuable; pure cloud-model or pure embedded specialists will see their individual competitiveness squeezed from both sides.

For the Consumer Market: Smart-feature stratification among two-wheeled EVs will accelerate, widening the experience gap between premium flagship models and basic commuter units; mid-tier OEMs that skip AI-ification will see their differentiation room further compressed.