PsyCloud

Getting Started

PsyCloud is a platform for building and running browser-based behavioral experiments. It handles the hard parts — reliable data capture, reproducible randomization, participant recruitment, hosting, and export — so you can focus on the experiment itself.

You can author a study visually in Studio, in code with the Python or R SDK, or by importing an existing experiment. Every path produces the same Project Bundle, compiles to the same runnable Study Bundle, and runs on the same runtime.

Choose your path

Not sure? Start with Studio.

If you're new, the Studio quickstart gets you to a complete, running study in a couple of minutes — and you can always export it to code later.

Pick by workflow

If you want to...Start here
Click together a study and preview it immediatelyStudio quickstart
Generate many study variants from data or scriptsPython quickstart or R quickstart
Serve, run, or compile an existing study locallyJavaScript runtime quickstart
Convert or maintain repo examplesCoffeeScript DSL quickstart
Bring an existing experiment into PsyCloudImporting & Migrating

The end-to-end workflow

However you author, a study moves through the same lifecycle:

  1. Author

    Define the structure, the trials, and what participants see — in Studio or code.

  2. Preview

    Run it against the real runtime with a simulated participant to debug logic and timing (Preview).

  3. Launch

    Validate, publish an immutable version, host a run, and recruit participants (Launch).

  4. Monitor & export

    Watch sessions arrive (Monitor), then inspect and export the data (Data).

What makes PsyCloud different

  • Studio-first, code-ready — a visual editor and two authoring SDKs that share one format, so you're never locked into one way of working.
  • Reliable capture — offline-first data collection that survives flaky connections.
  • Reproducible — deterministic, seedable randomization and explicit counterbalancing.
  • Auditable — strong provenance from design through to each recorded response.
  • Verifiable examples — canonical examples are checked across CoffeeScript, Python, R, and the bundled runtime contracts.

Next