Introducing SDK 2.0

We are on a mission to empower AI agents with robust, secure, and scalable virtual desktop infrastructure.

Today marks our first major update. We are excited to introduce SDK 2.0 for Python and TypeScript, an official REST API endpoint, four new protocol APIs (Browser, Code, File, and Env), subscription pricing for power users, and more.

TypeScript + REST API

Available today, spin up Scrapybara instances directly through the TypeScript SDK and REST API.

Browser Control

Scrapybara now supports programmatic browser control through Playwright. Empower your agents with direct HTML DOM input and manipulation (in addition to visual computer use controls) to better support web automation, data collection, and interaction testing. Learn more here.

Authentication Management

Authenticate automatically on popular websites like LinkedIn, X, GitHub, Amazon, etc. Our secure session management allows AI agents to maintain authenticated states across sessions, eliminating the need for frequent re-authentication while maintaining security standards.

Code Execution

Developers can now execute code directly within Scrapybara instances, and allow agents to launch and control entire Jupyter Notebooks (Python, Javascript, R). Learn more here.

Filesystem Control

We have introduced comprehensive filesystem management capabilities, giving AI agents full control over file operations. Learn more here.

Environment Variables

Our new environment variable interface allows for dynamic configuration of instance settings, making it easier to manage different deployment environments and application configurations. Learn more here.

Pricing

We are introducing three subscription tiers for power users that need more compute.

Self-Hosting

For organizations with specific security requirements or compliance needs, we now offer self-hosted deployment options. Get complete control over your infrastructure while maintaining all the benefits of our platform. Interested? Book a demo here.

Official and Community Cookbooks

To help you get started faster and share best practices, we are launching comprehensive cookbooks with recipes for common agent use cases. Our official examples cover essential implementations, while our community spotlight allows developers to share their own projects.

Submit your own projects for the community cookbook in our Discord using the #built-with-scrapybara channel.

Looking Forward

We are constantly optimizing performance across the Scrapybara platform, and working on supporting more operating systems, apps, and protocols to make it easier to useful agents. For developers interested in our upcoming releases, stay tuned for announcements in the coming weeks in our Discord and follow us on X.