Source: github.com/browser-use/browser-use/releases/tag/0.13.12
Browser Use has added OpenAI browser tools in version 0.13.12, backed by its Browser Use Actor. Developers import BrowserUse from browser_use.integrations.openai and install the browser-use[openai-tools] extra. The integration supports the 18 default browser tools and six optional actions.
Borrowed browser ownership is preserved, and Cloud browser cleanup is checked. URL and timeout settings carry over across OpenAI 2.x and 3.x model wrappers. OpenAI 3.x uses httpx2, so custom HTTP clients should use OpenAI's DefaultAsyncHttpxClient.
The integration does not yet support uploads, downloads, scoped reads, console or network diagnostics, or Bash. Browser Use reports 35 integration and model tests passed with two skipped, and 15 passed with one skipped on OpenAI 2.26. Live Cloud navigation and element-reference click tests also passed.
Why it matters if you build agents
Agents built on OpenAI's tooling can get browser control through this integration without writing their own browser tool wiring. Workflows that need file uploads, downloads, or Bash will still need other handling.
If you use a custom HTTP client with OpenAI 3.x, switch to OpenAI's DefaultAsyncHttpxClient before upgrading.
Get stories like this every weekday
One short email with the day's agent tooling news.