ScriptsApr 6, 2026·3 min read

Browser-Use Web UI — Visual AI Browser Automation

Gradio-based web interface for Browser-Use AI agent. Automate web browsing with visual feedback, persistent sessions, and HD recording. Supports 6+ LLM providers. 15,800+ stars, MIT.

TL;DR
Gradio-based web interface for Browser-Use AI agent with visual feedback, persistent sessions, and recording support.
§01

What it is

Browser-Use Web UI is a Gradio-based interface that wraps the Browser-Use AI agent. It lets you automate web browsing tasks through a visual dashboard rather than raw scripts. The agent navigates pages, fills forms, clicks buttons, and extracts data while you watch in real time.

This tool is built for QA engineers, growth hackers, and anyone who needs repeatable browser automation without writing Selenium or Playwright code from scratch. It supports six or more LLM providers for the underlying AI reasoning.

§02

How it saves time or tokens

Traditional browser automation requires writing and maintaining brittle selectors. Browser-Use replaces CSS selectors with natural language instructions interpreted by an LLM. The visual feedback loop catches failures immediately instead of after a full test suite run. Persistent sessions mean you do not re-authenticate on every run.

§03

How to use

  1. Clone the Browser-Use Web UI repository and install dependencies.
  2. Launch the Gradio server locally.
  3. Open the web UI, select your LLM provider, and type a natural language task.
# Clone and install
git clone https://github.com/browser-use/web-ui.git
cd web-ui
pip install -r requirements.txt

# Launch
python app.py
# Open http://localhost:7860 in your browser
§04

Example

Task: 'Go to Hacker News, find the top 3 posts, and save their titles and URLs to a CSV file.'

The agent will:
1. Navigate to news.ycombinator.com
2. Identify the top 3 posts by rank
3. Extract title and URL for each
4. Write results to output.csv
§05

Related on TokRepo

§06

Common pitfalls

  • Running without a visible browser window (headless mode) disables the visual feedback feature that makes debugging possible.
  • Some LLM providers have rate limits that cause the agent to stall mid-task; set appropriate retry and backoff parameters.
  • Persistent sessions store cookies locally; clear them between unrelated tasks to avoid cross-contamination of auth state.

Frequently Asked Questions

Which LLM providers does Browser-Use Web UI support?+

Browser-Use Web UI supports six or more LLM providers including OpenAI, Anthropic Claude, Google Gemini, and local models via Ollama. You select your provider and API key in the web UI settings panel before running a task.

Can I record the browser session as a video?+

Yes. The Web UI supports HD recording of browser sessions. Enable recording in the settings panel before starting a task. The output is saved as a video file you can review or share with your team.

Does it work with websites that require login?+

Yes. Persistent sessions retain cookies and auth tokens across runs. You can log in once during the first session, and subsequent runs will reuse the authenticated state until cookies expire or you clear them manually.

How does it compare to Playwright or Selenium?+

Browser-Use replaces CSS/XPath selectors with natural language instructions. You describe what you want done, and the AI figures out how to interact with the page. This eliminates selector maintenance but trades off precise control for convenience.

Is Browser-Use Web UI free to use?+

The Web UI itself is open source and free under the MIT license. However, you pay for the LLM API calls made during automation. Using a local model via Ollama eliminates API costs entirely.

Citations (3)
🙏

Source & Thanks

Created by Browser-Use. Licensed under MIT.

web-ui — ⭐ 15,800+

Discussion

Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.

Related Assets