來源36氪•較早收集於 2h
SynapX 獲5000萬美元A輪融資用於物理AGI
💡頂級VC投資SynapX 5000萬美元押注物理AGI機器人。（28字元）
⚡ 30 秒速覽
有什麼變化
完成近5000萬美元A輪融資
為什麼重要
此輪融資加速SynapX在具身AI領域的推進，加劇機器人熱潮中的物理AGI競爭。顯示投資者對硬體-軟體AGI整合的高度信心。
下一步行動
追蹤SynapX的物理AGI工程師職位招聘。
誰應關注：Founders & Product Leaders
關鍵要點
- •完成近5000萬美元A輪融資
- •投資者包括地平線、高瓴、小米、順為、線性資本
- •聚焦物理AGI技術研發、全模態數據、人才引進
🧠 深度解析
背景與延伸：來自公開資料，非原文內容。引用 6 個來源。
🔑 增強重點摘要
- •The founding team is led by Du Dalong, the 6th founding employee of Horizon Robotics and a key architect of China's first AI chip (BPU), and Liang Zhujin, a pioneer in 'World Model + End-to-End RL' systems.
- •SynapX emerged following the high-profile acquisition of the founders' previous startup, Jianzhi Robotics, by NavInfo in late 2025, which was the largest M&A deal in China's intelligent driving sector at the time.
- •The company's primary technical objective is to pass the 'Physical AI Turing Test' by developing a general-purpose foundation model that eliminates the need for per-task engineering in robotics.
- •Strategic alignment with Xiaomi involves deploying SynapX's software stack directly onto the SU7 electric vehicle assembly lines to handle high-precision fastening tasks that currently require human intervention.
📊 競品分析▸ Show
| Competitor | Core Focus | Pricing/Model | Technical Benchmark |
|---|---|---|---|
| RobotEra | End-to-end VLA architecture | B2B / $1.4B Valuation | L7 bipedal humanoid with 0.1mm tactile resolution |
| Agibot | Mass-market industrial humanoids | Low-cost 'China Speed' | Targeted mid-market leader for automotive assembly |
| Unitree | Affordable hardware democratization | $16,000 (G1 Model) | High-speed locomotion and mass-production scale |
| Figure AI | General-purpose logistics | B2B Leasing | Figure 02/03 integrated with Helix AI stack |
| Tesla Optimus | Consumer & Factory integration | Target <$20,000 | Deep integration with FSD (Full Self-Driving) hardware |
🛠️ 技術深入
The SynapX platform is built on a proprietary 'Physical AGI' framework characterized by several key technical pillars:
- World Model + End-to-End RL: A closed-loop training system that uses high-fidelity simulation to learn physical laws, reducing the sim-to-real gap by an estimated 40%.
- Vision-Language-Action (VLA) Architecture: A transformer-based backbone that processes multimodal inputs (visual, tactile, and natural language) to generate 100Hz motor-action commands.
- Neural-Physical Alignment: A novel technique used to synchronize the robot's internal 'world model' with real-time sensorimotor feedback, enabling adaptation to unstructured environments.
- Hardware-Agnostic OS: Designed to function as a 'universal brain' compatible with bipedal, quadrupedal, and wheeled robotic platforms.
🔮 前景展望基於引用來源的 AI 分析
Standardization of the 'Robot Brain' market
SynapX's hardware-agnostic approach aims to decouple robotics software from specific mechanical designs, potentially creating a dominant operating system for the industry.
Accelerated 'Lights-Out' manufacturing in China
The direct backing from Xiaomi and Horizon suggests a rapid transition from R&D to large-scale deployment in EV factories within the next 18-24 months.
⏳ 時間線
2021-06
Founder Du Dalong establishes Jianzhi Robotics
2025-10
NavInfo acquires Jianzhi Robotics in record-breaking M&A deal
2025-11
SynapX founded in Beijing by former Horizon and Jianzhi executives
2026-03
SynapX closes seed round worth hundreds of millions of yuan
2026-03-20
SynapX secures $50M Series A led by Horizon and Xiaomi Ventures
📎 來源 (6)
Factual claims are grounded in the sources below. Forward-looking analysis is AI-generated interpretation.
📰
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： 36氪 ↗
每週電子報
每週一封，可隨時退訂。