How is pricing calculated?

Pricing is based on the hourly rate of operators in your selected region. You only pay for the time operators spend controlling your robots.

What robots are supported?

We support 9+ robot arms including SO-100, Koch, Franka FR3, Franka Panda, WidowX 250, ViperX 300, ViperX 300s, and ALOHA.

What is the latency for teleoperation?

Our platform provides ultra-low latency of 30-70ms worldwide, enabling real-time responsive robot control.

SO-100 / SO-ARM100

Software for the SO-100 robot arm

Record imitation learning demonstrations, train a policy in the cloud, and run it back on your own arm. One app, no Python environment to maintain.

Download the client Getting started guide

What the SO-100 is

The SO-100, also sold as the SO-ARM100, is an open-hardware five-degree-of-freedom arm with a gripper, built around STS3215 serial bus servos. It was designed as an affordable way to collect manipulation data, and it has become the default arm for people learning robot policies without a research budget.

It has no force-torque sensing and modest repeatability, which makes classical model-based control a poor fit. That is precisely why it pairs well with imitation learning: you demonstrate the task instead of modelling it.

5 DoF plus gripper

STS3215 serial bus servos

Low cost

Open hardware from TheRobotStudio

Camera driven

Wrist and scene views feed the policy

LeRobot native

Records straight into the standard format

Start here

Four guides that take you from an unopened box to a policy running on the arm.

SO-100 getting started: from the box to the first movement

Assemble, wire and calibrate an SO-100 robot arm, install the AY-Robots client, and move the arm for the first time. No prior robotics experience required.

15 min · beginner

Record your first imitation learning dataset

Record fifty usable demonstrations with an SO-100 robot arm, review them, and upload a LeRobot dataset ready for training.

20 min · beginner

Train your first policy

Start a training run on an uploaded SO-100 dataset, choose a policy architecture, read the loss curve honestly, and pick the right checkpoint.

15 min · intermediate

Run your first policy on the arm

Deploy a trained policy to a real SO-100, evaluate it honestly, diagnose the common failure modes, and decide what data to record next.

15 min · intermediate

Go deeper

Reference material on the parts people get stuck on.

Recording imitation learning data with the SO-100

How to record clean imitation learning demonstrations with an SO-100 robot arm: episode structure, how many demos you actually need, and the mistakes that quietly ruin a dataset.

The SO-100 and the LeRobot dataset format

How SO-100 recordings map onto the LeRobot dataset format, what changed between v2.1 and v3.0, and which format the common policy trainers actually accept.

Imitation learning on the SO-100, from demonstrations to a policy

What imitation learning is, why it suits a low-cost arm like the SO-100, which policy architectures are worth your time, and what to expect at each stage.

Ready when your arm is

Download the client for macOS, Windows or Linux and record your first episode today.

Download the client