Freedom251 › AI Browser Agents
Auto Browser

What Auto Browser is
Auto Browser is a Chrome side-panel AI agent that reads webpages and carries out tasks described in plain English. It can inspect a page’s accessibility tree, click, type, scroll, navigate, fill forms and handle research tasks. On pages that expose WebMCP tools, it can use those tools instead of interacting through the page. It supports Chrome Built-in AI (Gemini Nano), Google Gemini, OpenRouter and local OpenAI-compatible endpoints such as Ollama and LM Studio. You can switch providers mid-conversation while keeping the conversation history. The extension is free to install. Google Gemini and OpenRouter charge per request when you use your own API key. The maker says it has no extension backend and does not receive or store prompts, browsing activity, conversation history or API keys. With a cloud provider, prompts and relevant page context go directly from the browser to that provider; the maker says Gemini Nano and local LLM options send no data from the device through the extension. Basic use requires Chrome 126 or newer. Gemini Nano has additional browser, operating system, memory and disk requirements. Write actions require approval, arbitrary JavaScript always prompts, and banking, government and healthcare origins are blocked by default.
Who it suits
Auto Browser suits Chrome users who want an AI agent to read pages, fill forms or handle research tasks, with a choice of cloud or local AI providers. It also fits users who want approval steps for write actions and page interaction.
What it does well
- Reads accessibility trees and interacts with pages.
- Uses WebMCP tools when pages expose them.
- Supports cloud, built-in and local AI providers.
- Provider switching keeps conversation history.
- Write actions require approval.
- Free to install.
What to know first
- Requires Chrome 126 or newer.
- Gemini Nano requires Chrome 138 or newer and substantial device resources.
- Cloud AI providers receive prompts and relevant page context.
- Terms prohibit automated pipelines, bots, crawlers and commercial services beyond ordinary individual use.
Limits
The extension is free to install, but Google Gemini and OpenRouter charge per request when used with your API key. Gemini Nano requires Chrome 138 or newer, macOS 13 or newer, Windows 10 or newer, Linux or ChromeOS, 16 GB RAM and 22 GB free disk space.
Plans
| Plan | Billed and limits | Price |
|---|---|---|
| Free | Extension is free to install · Google Gemini and OpenRouter charge per request using your API key | Free |
As autobrowser.dev lists them, read 3 Oct 2026. The maker's own page has the final word.
Getting started
- Open the Auto Browser website.
- Use Chrome 126 or newer.
- Install the free extension from the Chrome Web Store.
- Choose a supported AI provider; Google Gemini and OpenRouter use your API key and charge per request.
- For Gemini Nano, meet its Chrome, operating system, memory and disk requirements.
Verdict
Choose Auto Browser if you want a free Chrome extension for page tasks and the ability to select among built-in, cloud or local AI. Look elsewhere if you need a browser other than Chrome or want to use it for automated pipelines, bots, crawlers or commercial services beyond ordinary individual use.
Questions
How much does Auto Browser cost?
The extension is free to install, listed as free. Google Gemini and OpenRouter charge per request when used with your API key.
Which browsers does it support?
Auto Browser requires Chrome 126 or newer for basic use.
Can I use a local AI provider?
Yes. It supports local OpenAI-compatible endpoints such as Ollama and LM Studio. The maker says local LLM options send no data from the device through the extension.
Does Auto Browser store my prompts or browsing activity?
The maker says it operates no extension backend and does not receive or store prompts, browsing activity, conversation history or API keys. When you select a cloud provider, prompts and relevant page context go directly from the browser to that provider.
Can Auto Browser perform actions without approval?
Write actions require approval, and arbitrary JavaScript always prompts. Banking, government and healthcare origins are blocked by default.
What are the terms of use?
The terms prohibit using the service to operate an automated pipeline, bot, crawler or commercial service beyond ordinary individual use.
Facts from Auto Browser's own pages
- AI providers
- It supports Chrome Built-in AI (Gemini Nano), Google Gemini, OpenRouter, and local OpenAI-compatible endpoints such as Ollama and LM Studio. · autobrowser.dev, 3 Oct 2026
- Attachments
- Users can attach images and audio files of up to 10 MB each in the side panel. · autobrowser.dev, 3 Oct 2026
- Cloud data handling
- When a cloud AI provider is selected, prompts and relevant page context are sent directly from the browser to that provider. · autobrowser.dev, 3 Oct 2026
- Integrations
- Pages that expose tools using WebMCP can let Auto Browser use those tools instead of interacting through the DOM. · autobrowser.dev, 3 Oct 2026
- Local AI
- The maker says Gemini Nano and local LLM options send no data from the device through the extension. · autobrowser.dev, 3 Oct 2026
- Notable restrictions
- The terms prohibit using the service to operate an automated pipeline, bot, crawler, or commercial service beyond ordinary individual use. · autobrowser.dev, 3 Oct 2026
- On-device requirements
- Gemini Nano requires Chrome 138 or newer, macOS 13 or newer, Windows 10 or newer, Linux or ChromeOS, 16 GB RAM, and 22 GB free disk. · autobrowser.dev, 3 Oct 2026
- Page interaction
- It reads the page accessibility tree and can click, type, scroll, navigate, and use page-provided WebMCP tools. · autobrowser.dev, 3 Oct 2026
- Privacy
- The maker says it operates no extension backend and does not receive or store prompts, browsing activity, conversation history, or API keys. · autobrowser.dev, 3 Oct 2026
- Provider switching
- Users can switch AI providers mid-conversation while retaining conversation history. · autobrowser.dev, 3 Oct 2026
- Safety controls
- Write actions require approval, arbitrary JavaScript always prompts, and banking, government, and healthcare origins are blocked by default. · autobrowser.dev, 3 Oct 2026
- Support
- The maker directs users to the Chrome Web Store listing’s support channel for help with blocked sites or other issues. · autobrowser.dev, 3 Oct 2026
- Supported browser
- Auto Browser requires Chrome 126 or newer for basic use. · autobrowser.dev, 3 Oct 2026
- What it does
- Auto Browser is a Chrome side-panel AI agent that reads webpages and carries out tasks described in plain English. · autobrowser.dev, 3 Oct 2026
Where Auto Browser stands on Freedom251
- AI Browser Agents#6 of 22
- Browser Agents#6 of 18
