SUMMARY

 

RAISE A1 is AgiBot’s first‑generation general‑purpose humanoid robot, launched in August 2023 and mass‑produced beginning in 2024.

 

Standing 175 cm and weighing 55 kg, it delivers 49+ DOF, 7 km/h walking speed, 5 kg per‑arm payload, and 200 TOPS onboard AI compute.

 

It is the foundational platform that enabled AgiBot to scale to 5,000+ humanoids shipped by late 2025, powering deployments in manufacturing, logistics, warehousing, research, and service robotics.

 

EDITORIAL

 

RAISE A1 is the robot that put AgiBot on the global map.

 

It introduced the company’s EI‑Brain architecture — a four‑layer embodied‑intelligence stack (Cloud Superbrain → Main Brain → Sub‑brain → Brainstem) — and WorkGPT, AgiBot’s proprietary multimodal LLM for mission‑level planning.

 

Mechanically, A1 pioneered PowerFlow quasi‑direct‑drive joints, delivering ~350 N·m peak torque at only 1.6 kg per joint, with liquid cooling, dual encoders, and <10:1 planetary reducers.

 

Its SkillHand system — 12 active + 5 passive DOF with visual fingertip sensors — remains one of the most advanced dexterous hands in its class.

 

RAISE A1 is not a demo robot; it has been deployed in real industrial pilots, including bolt tightening, vehicle inspection, material induction, and warehouse logistics.

 

SPECIFICATIONS

 

Dimensions & Build

 

• Height: 175 cm

• Weight: 55 kg

• Frame: Aluminum + composite structural materials

• Knee design: Reverse‑jointed for stability

 

Mobility

 

• Walking speed: 7 km/h

• Locomotion: Bipedal, reverse‑knee architecture

• Navigation: LiDAR‑based SLAM + visual odometry

 

Degrees of Freedom

 

• 49+ DOF total

• Hands: 12 active + 5 passive DOF

 

Actuation & Manipulation

 

• Joint system: PowerFlow quasi‑direct‑drive

• Peak torque: ~350 N·m per joint

• Arm payload: 5 kg per arm

• Deadlift capacity: 80 kg

• Hand sensors: Visual‑based fingertip tactile sensors

 

Perception & Sensors

 

• RGB‑D cameras

• LiDAR

• Microphone array

• IMU + gyroscope

• Force/torque sensors

• Proximity sensors

 

Compute & AI

 

• Onboard compute: 200 TOPS

• AI stack:

• EI‑Brain (Cloud → Main → Sub → Brainstem)

• WorkGPT multimodal LLM

• Hybrid model‑based + learning‑based control

• Software: Proprietary OS + ROS2 + Python + C++

 

Battery & Power

 

• Runtime: ~120 minutes

• Cooling: Liquid‑cooling circulation system

 

Connectivity

 

• Wi‑Fi (2.4/5 GHz)

• Ethernet

• Bluetooth

• USB / HDMI

 

Applications

 

• Manufacturing

• Logistics & warehousing

• Precision assembly

• Inspection

• Service robotics

• Research & education

 

Pricing

 

• $30,000 USD (2026 market listing)

 

WHAT IS NOT PUBLISHED

 

• IP rating

• Battery capacity (Wh)

• Exact torque per joint for all axes

• Thermal limits

• Long‑term durability metrics

 

Image: @AIStackTeam

 

COMPANY SOURCE

 

https://youtube.com/shorts/-FSBDsKeb1s?si=uZtQQOVDIsI2XvDy

 

KEY NEWS & COVERAGE

 

https://youtube.com/shorts/-FSBDsKeb1s?si=uZtQQOVDIsI2XvDy