Xpeng 2026 年將再聘 8,000 人押注物理 AI

💡Xpeng hires 8K, masses robots/Robotaxis in 2026—must-read for physical AI strategists.
⚡ 30-Second TL;DR
有什麼變化
2026 年計畫再聘 8,000 名員工
為什麼重要
Xpeng 的招聘激增與生產計畫加劇具身 AI 自動駕駛與機器人領域競爭。它定位公司為 EV 相關 AI 硬體的全球化領導者。
下一步行動
Benchmark Xpeng's Robotaxi physical AI stack against your autonomy models for globalization insights.
關鍵要點
- •2026 年計畫再聘 8,000 名員工
- •押注「物理 AI + 全球化」策略
- •同時量產機器人與飛行汽車
- •Robotaxi 作為 2026 年核心計畫
🧠 深度解析
背景與延伸:來自公開資料,非原文內容。引用 7 個來源。
🔑 增強重點摘要
- •XPeng plans to hire 8,000 additional staff in 2026 to support its 'physical AI + globalization' strategy, focusing on commercialization of embodied AI in mobility, robots, and flying cars.[1]
- •Chairman He Xiaopeng emphasized 2026 as the crucial year for physical AI commercialization, with mass production of L3-level humanoid robots for industrial use and breakthroughs expected in 12-18 months.[1]
- •XPeng is launching three mass-produced Robotaxi models in 2026 with trial operations, powered by second-generation VLA model (72 billion parameters) achieving L4 autonomy via end-to-end visual-to-action generation and four Turing chips (3,000 TOPS).[2][3]
- •New IRON humanoid robot for mass production by late 2026, leveraging autonomous driving tech; AeroHT eVTOL flying car rollout in Q3 2026 for under 2 million RMB, including Land Aircraft Carrier.[2][4]
- •XPeng merged autonomous driving and smart cockpit departments to form a General Intelligence Center, extending VLA tech to robotics and flying cars; ahead of industry by 5 years per partners.[1][2]
📊 競品分析▸ Show
| Feature | XPeng | Competitors (e.g., Tesla, others) |
|---|---|---|
| Robotaxi Models (2026) | 3 models, L4 VLA, 3,000 TOPS | Tesla: Cybercab planned; Pony.ai testing in China |
| Humanoid Robot | IRON, L3 industrial mass-prod | Tesla Optimus; Figure AI prototypes |
| Flying Car/eVTOL | AeroHT Q3 2026, <2M RMB | EHang, Lilium focusing on air taxis |
| AI Chip | Turing 700 TOPS/chip | Tesla HW5; Nvidia Orin |
| Hiring/Expansion | 8K staff for physical AI | Tesla: Aggressive robotics hiring |
🛠️ 技術深入
- Second-gen VLA Model: 72B parameters, end-to-end from visual signals to actions (bypasses language), trained on real-world driving data for L4 in all scenarios, no geographic limits, mass-produced first for physical world.[2]
- Turing AI Chip: 700 TOPS per chip, deployed in models; Robotaxi/GX uses 4 chips for 3,000 TOPS effective compute, powers smart driving/cockpit.[3]
- SEPA 3.0 Architecture: For GX SUV, includes steer-by-wire, rear-wheel steering; enables L4 testing.[3]
- IRON Humanoid: Advanced balance/posture mimicking humans, mechanical systems for uneven surfaces; applies VLA/autonomous driving tech for industrial L3 mass production.[1][4]
- General Intelligence Center: Merges driving/cockpit depts, unifies AI infrastructure for vehicles, robots, flying cars.[2]
🔮 前景展望AI analysis grounded in cited sources
XPeng's aggressive 8K hiring and 2026 mass production across Robotaxis, robots, and eVTOLs positions it as a leader in physical AI mobility, potentially disrupting global EV/autonomous markets by integrating full-stack AI (hardware/software/data) and expanding via globalization, challenging Tesla in embodied AI while leveraging China-US AI race dynamics.
⏳ 時間線
📎 來源 (7)
Factual claims are grounded in the sources below. Forward-looking analysis is AI-generated interpretation.
- eu.36kr.com — 3662222911710084
- autonews.gasgoo.com — Xpeng Merges Autonomous Driving Smart Cockpit Departments to Establish General Intelligence Center 2019041782581329921
- cnevpost.com — Xpeng Begins Testing Level 4 Autonomous Driving on Gx Suv
- time.com — China US AI Race
- marketscreener.com — Xpeng Makes First Appearance at World Governments Summit Highlights AI Driven Future of Mobility Ce7e5ddadf8cfe23
- euronews.com — World Governments Summit Flying Cars and Robots Move From Science Fiction to Reality
- cleantechnica.com — Xpeng Isnt Just Entering Asean Its Assembling an Operating System
AI 週報
閱讀本週精選 AI 大事摘要 →
👉相關動態
AI 策展新聞聚合。所有內容版權歸原始發布者所有。
原始來源: Pandaily ↗
每週 AI 簡報
每週一封,可隨時退訂。