AI Agent News/stories/browser-use-0-13-14-adds-openai-browser-tools-and-cloud-sdk-v4-imports.md

Browsers

browser-use 0.13.14 adds OpenAI browser tools and Cloud SDK V4 imports

browser-use has released version 0.13.14 of its Python agent package, adding OpenAI browser tools and a shared Bash helper. The update also improves page observations and browser recovery, and adds support for Browser Use Cloud SDK V4 imports.

Source: github.com/browser-use/browser-use/releases/tag/0.13.14

The release adds OpenAI browser tools and a shared Bash helper. It also improves bounded page observations, form state, file access, and recovery from browser failures. Fixes cover temporary empty Cloud tab URLs, alignment of screenshots, crops, and clicks on scaled displays, and cleanup after failed browser reconnect initialization.

Missing-key errors for Cloud browsers now give clearer messages. The aiohttp dependency moves from 3.14.3 to 3.14.5, and the Cloud guidance and saved-task replay examples have been refreshed. The Python agent and its existing models are unchanged, and the Pi migration is handled separately.

According to the release notes, GitHub checks pass on the exact release source, and the wheel and source package build successfully. The installed wheel runs browser-use --help, and all five CLI entries pass validation. Install with uv add browser-use==0.13.14 or pip install browser-use==0.13.14, though PyPI publication waits for approval from another release reviewer.

Why it matters if you build agents

Developers who run Cloud browsers get fixes for empty tab URLs and reconnect cleanup, along with clearer missing-key errors. Teams that depend on screenshot coordinates on scaled displays should check that clicks now line up as expected. Until PyPI publication completes, the pinned install command may not yet resolve to the new version.

Get stories like this every weekday

One short email with the day's agent tooling news.

Subscribe