Next robot →

AheadForm Origin F1 / M1 / Elf Xuan

AheadForm (China) · 2025

Quote onlyNo published price (October 2026 check). Origin F1 and Elf Xuan are ordered through an inquiry and reservation form on aheadform.com, and the Origin M1 page still says coming soon. Media figures of about $150,000 for the most advanced Origin M1 (unconfirmed) and $50,000 – $150,000 for Origin F1 (unconfirmed) do not come from AheadForm. Year is the Origin M1 reveal in 2025.
AvailableNot programmableLifelike facePupil camerasBrushless micro-motorsDesktop bust
Weeks+Shipping: Quote. Sold to businesses through an inquiry form on aheadform.com. Lead times and US delivery terms are not published. AheadForm inquiry ↗
Remote control: Not possible with public tools. Any development access goes through AheadForm's Origin F1 development interface, which has no public documentation.

Overview

Hyper-realistic robot faces from a startup founded in December 2024 by Yuhang Hu, a Columbia University PhD from the Creative Machines Lab. The male Origin M1 bust went viral in 2025 for blinking, nodding and micro-expressions driven by up to 25 brushless micro-motors (unconfirmed) under silicone skin, with cameras in its pupils. Origin F1 is the standardized desktop version with 26 degrees of freedom across head, eyes and face, skin that shows veins and small muscle shifts, and AheadForm's own multimodal AI, and it drew crowds at WAIC in Shanghai in July 2026. The Elf series, led by the elf-eared Elf Xuan with 30+ facial degrees of freedom, is a stage performer that sang at a February 2026 robot gala. It is a talking face like Furhat and rivals Moya for realism, but there is no public price, spec sheet or SDK.

Specifications

CategoryCompanion robots · Desk & social robots
Released2025
CountryChina
AvailabilityAvailable
ProgrammabilityLimited SDK
Built-in autonomyReactive · Responds to its surroundings without a map, such as avoiding obstacles or following a person
LLM supportBuilt-in LLM chat · AheadForm's own CharacterMind multimodal model combines voice, vision, context and emotion so the robot chooses its gaze, smiles and nods while it talks. AheadForm (Chinese site) ↗
MuJoCo simulationNone found · Nothing found.
NVIDIA Isaac simulationNone found · Nothing found.
Gazebo simulationNone found · Nothing found.
Best forAges 18+ (estimated)
Vision sensorsCamera
ActuatorsBrushless / direct-drive motors
AudioSpeaker, Microphone
Compute typeUnknown
Remote controlNot programmable
SizeDesktop head-and-shoulders bust (Origin F1 and M1). Elf Xuan is a seated full figure, and Elf Xuan 2.0 adds a moving upper body. Dimensions not published
DriveStationary. Elf Xuan 2.0 also moves its waist and arms
ComputeNot published
VisionCameras embedded in the pupils for eye contact and recognition
LidarNone
SensorsPupil cameras, microphones and speakers, plus conversation memory
RuntimeNot published
SpeedN/A
SDKNo public SDK. The official Chinese Origin F1 page says it ships with a stable system and development interface for secondary development and algorithm tuning

Ratings & reviews

Similar robots

Image credits