Skip to main content

​
New features

​
Volume CRUD operations in SDKs

You can now create, list, and delete volumes directly from the JavaScript and Python SDKs. Previously, volume management required the dashboard or API. This makes it easier to programmatically manage persistent storage for your sandboxes.

​
Gzip content encoding for file operations

File uploads and downloads now support gzip compression. This reduces bandwidth usage and speeds up transfers, especially for large text-based files. The SDKs handle compression automatically when enabled.

​
CLI info command

A new e2b sandbox info command lets you inspect a running sandbox’s details, including its lifecycle state and network configuration. See the CLI docs for usage.

​
Updates

Pausing a sandbox is now significantly faster. The platform uses copy-on-write filesystem reflinks during pause, which avoids copying unchanged data. You should notice reduced pause times, especially for sandboxes with large filesystems.

​
Increased volume upload timeout

Volume file uploads now have a 1-hour timeout (previously shorter), so you can upload larger datasets to volumes without running into timeout errors.

​
Binary file uploads use application/octet-stream

Sandbox file uploads now use the application/octet-stream content type, improving compatibility with binary file transfers.

​
Firecracker v1.12

The underlying VM runtime has been upgraded to Firecracker v1.12, bringing performance improvements and better resource management for sandboxes.

​
Bug fixes

  • Snapshot deletion on CLI create: The e2b sandbox create command no longer inadvertently deletes snapshots when shutting down a sandbox.
  • Connection config propagation: Default connection configuration is now correctly applied across all SDK operations.
  • Better error messages: Sandbox-not-found and file-not-found errors are now distinguished, so you get clearer feedback when something goes wrong.
  • Paused sandbox template names: Template names now resolve correctly for paused sandboxes in the API and dashboard.
  • Auto-resume reliability: Improved state handling during auto-resume reduces edge cases where sandboxes could fail to resume properly.