Skip to content
Qualia Docs

Introduction

What is Qualia?

Section titled “What is Qualia?”

Qualia is a cloud platform for robot data. Upload robot recordings, curate their episodes, and save immutable, reproducible dataset versions that you can download.

Qualia takes three kinds of input:

  • MCAP ROS 2 recordings, from a teleoperated arm or similar rig
  • UMI handheld recordings, which convert as they upload when you pass --rig
  • LeRobot datasets, v2.1 or v3

Getting Started

Section titled “Getting Started”

  1. Install the CLI. The qualia CLI wraps the Python SDK and installs it for you. The commands below have the exact version for this deployment; see Installation.

    Terminal window
    pip install 'qualia-cli==0.3.0rc11'
    

    # or with uv
    uv tool install 'qualia-cli==0.3.0rc11' --with 'qualia-sdk==0.8.0rc11'

    This deployment runs ahead of the public release, so it needs the matching pre-release wheel — 0.3.0rc11. Pin it rather than upgrading: the upload protocol is negotiated exactly, so the client and the endpoint it talks to have to be the same generation. A plain pip install qualia-cli resolves to the current public release and will be turned away by this deployment.

  2. Create an API key. In the Qualia dashboard, open Settings → Account → API keys → New key, and choose the whole organization or one workspace. Copy it when it appears; it is shown once. Viewers need a workspace owner to grant write access before uploading data.

  3. Log in. Use the API origin for your deployment:

    Terminal window
    qualia auth login --base-url https://dev-api.qualiastudios.dev --api-key <YOUR_KEY>

    Replace <YOUR_KEY> with your API key. For an interactive prompt that hides the key, omit --api-key. Scripts and CI can set QUALIA_API_KEY and QUALIA_BASE_URL instead; set both together. The key names its own account. See Authentication for the API origin and where the key is saved.

  4. Upload. Upload a local directory:

    Terminal window
    qualia dataset upload <dir> <name> --source-format mcap_ros2

    Replace <dir> with your recording directory and <name> with a dataset name. For LeRobot v3, use --source-format lerobot_v3. See Upload for other formats.

The Data Workflow

Section titled “The Data Workflow”
  1. Upload an MCAP ROS 2 or UMI recording, or a LeRobot dataset, with qualia dataset upload.
  2. Add a robot and a task: add the robot from its URDF, set it as the dataset’s target robot, and link the dataset to a task.
  3. Curate the episodes: keep or reject them, trim them, mark subtasks and issues, and save versions.
  4. Download a saved version as a LeRobot directory.

Resources

Section titled “Resources”