Microduck
Made to move · Ready to learn
A 25 cm open-source biped you train yourself with reinforcement learning. Playable out of the box.
Roll the Tape
The launch film · sound on
Trained in Sim, Deployed on the Real Robot
Every behaviour starts as a neural policy in physics simulation, then runs onboard at 50 Hz. Fun out of the box. Yours to retrain.
Teach It New Tricks
Every behaviour is a policy you can retrain on your own machine.
01 — Train in Simulation
Behaviours are learned in physics sim, on your machine or in the cloud.
02 — Deploy on the Robot
One step from simulation to the real thing.
03 — Refine
Tune, re-train, re-deploy. Iterate as much as you like.
04 — Publish the Policy
Share your new behaviour with the community!
Skills in the Box
Walk
Smooth velocity-tracking gait.
Sit & Stand
Sits down, holds the pose, stands back up.
Kick
A quick boot, then straight back to walking.
Grab
Dips the beak, scoops, pops back upright.
Roller Skating
Cruises on skates when equipped.
Get Back Up
From flat on its back to standing, all by itself.
One robot, four colourways
Choose Your Colour
Every Microduck ships in one of four colourways. Same robot, same brains underneath — pick the shell that best fits you.
Sunny Yellow
Ocean Blue
Bubble Pink
Mint Green
In the Wild
The real robot in real places — on desks, on the pitch, out at golden hour.
Pick Your Pack
The robot is everything you need on day one. The packs add play gear and spare parts.
Tech Specs
15
Motors
25 cm
Tall
800 g
Lightweight
Camera
Plus LiDAR & dual IMUs
7
Trained moves in the box
50 Hz
Onboard policy loop
Built in the Open
The SDK, the simulation and the full RL training stack are open source. What the robot runs is what you can read, fork and retrain.
Apache-2.0 — the whole software stack, permissively licensed.
Join the Flock
Builds on show, policies to swap, help when a leg does something strange.
Pre-orders Are Open Now
Introductory price, before taxes and shipping.