Pinecone Pie - The first AI-native edge-side intelligent development board launched by Facewall Intelligence

What is Pinea Pi? Pinea Pi is the first AI-native edge-side intelligent development board launched by Facewall Intelligence. Built on the NVIDIA Jetson AGX Orin platform, it features built-in multimodal components such as a camera and microphone. Pinea Pi supports direct-drive development using natural language, out-of-the-box MiniCPM multimodal large models, and full-chain offline operation, allowing for the development of intelligent hardware without coding. It is suitable for scenarios such as personal assistants, embodied intelligence, and programming education, and is expected to be available in mid-2026. Pinea Pi's main functions include Agent...

Pinecone Pie - The first AI-native edge-side intelligent development board launched by Facewall Intelligence

What is pinecone pie?

Pinea Pi is the first AI native end-to-side intelligent development board launched by Facewall Intelligence. It is based on NVIDIA Jetson AGX Orin and has built-in multi-modal components such as cameras and microphones. Pinecone Pie supports natural language direct-drive development, and the MiniCPM multi-modal large model can be used out of the box to realize full-link offline operation. It can develop intelligent hardware without coding. It is suitable for personal assistants, embodied intelligence, programming education and other scenarios. It is expected to be launched in mid-2026.

Main functions of Pinecone Pie

  • Agent native : Support developers to directly drive hardware development through natural language, convert intentions directly into execution actions, achieve intelligent interaction without writing code, and significantly reduce the technical threshold of end-side intelligent development.
  • Multimodal native : Built-in wall-facing intelligent self-developed MiniCPM-V The multi-modal large model and the MiniCPM-o full-modal large model use visual, auditory, language and other perceptual capabilities out of the box, allowing the hardware to have a complete interactive experience of “seeing, listening, and speaking”.
  • End-to-end native : Supports full-link offline operation of large models, and the data is completely retained on the local side, ensuring user privacy is safe and controllable, the user experience is stable and smooth, and there is no need to continuously pay token fees for model calls.

Specifications of Pine Cone Pie

  • Core computing power : NVIDIA Jetson AGX Orin 64GB
  • interactive components : Camera + Microphone Array + Speaker
  • storage expansion :M.2 NVMe SSD
  • network connection :WiFi 6 + Bluetooth 5.2 + Gigabit network port
  • Interface configuration : USB-A × 2, USB-C × 2, HDMI
  • Cooling system : Active heat dissipation, no frequency reduction under full load

Pine cone pie price configuration

Pine Cone Pie is planned to be launched this year, and the price will be based on the official launch price.

Where to buy Pine Cone Pie

Pine Cone Pie is not yet on the market, please pay attention to the official website for information.

Pineapi official website address: https://pineapi.modelbest.cn/.

Application scenarios of Pinecone Pie

  • Offline multi-modal personal knowledge assistant : Create a privatized intelligent assistant based on a local large model that supports document understanding, voice interaction, and image recognition, and data does not come out to ensure privacy and security.
  • Embodied intelligence development : Provides end-side “brains” for robots, smart cars and other hardware to achieve natural language control, environmental perception and independent decision-making, and lower the threshold for robot development.
  • AI programming educational teaching aids : As a teaching platform, it helps students understand the principles of multi-modal AI and quickly get started with intelligent hardware development through natural language interaction and visual programming.
  • Smart home hub : Offline control of smart devices, supporting voice commands, visual recognition and other interactive methods, to create a localized smart living system that does not require the Internet.
  • edge computing node : Use Jetson AGX Orin computing power to complete AI inference locally, suitable for scenarios with low latency and high privacy requirements such as industrial quality inspection and security monitoring. ©