Current location: Home> Ai News

Free open source AI automation tool Nanobrowser: local deployment, support for multiple LLM, new choice for web operation

Author: LoRA Time: 12 Mar 2025 431

Still hurting for the hundreds of dollars a month of OpenAI Operator subscription fees? The emergence of Nanobrowser is undoubtedly a shot of heart. It is a completely free open source tool that requires no subscription fees and completely frees you from financial constraints. All you need to do is install this extension and configure your own LLM API key to enjoy the top web automation capabilities right away. This "bring your own dry food" method is not only economical but also makes you know the cost well and truly "I am the one who is in charge of my AI."

QQ_1741744451353.png

Local deployment, security Max

In this era of increasingly important data security, Nanobrowser puts user privacy first. Unlike those services that run in the cloud, all operations of Nanobrowser are performed in your local browser . This means that your API keys and all operational data are firmly in your hands, without worrying about the risk of sensitive information leakage. This "local priority" design concept undoubtedly provides peace of mind for users who pay attention to privacy.

None of the three AI giants can be missing

Nanobrowser shows amazing openness in support of LLM (large language model). It currently supports the API of three mainstream AI models including OpenAI, Anthropic (Claude) and Google (Gemini). You can freely choose the most suitable LLM based on different task requirements and cost considerations, and even configure different models for different AI agents. In the future, Nanobrowser will also actively expand support for more LLMs to truly achieve "a hundred flowers bloom and each one takes what they need."

Visualization operations, novices can become experts

Have you ever been dissuaded from the operation interface of complex RPA (robot process automation) tools? Nanobrowser provides you with a friendly operation interface with its intuitive interactive sidebar . You can use simple instructions to automatically complete various web tasks like chatting with AI. Whether it is to capture news headlines , search for popular projects on GitHub , or compare prices on e-commerce platforms , Nanobrowser can understand and execute efficiently in just one sentence. Its multi-agent system is a highlight. Different AI agents can work together to complete complex web workflows, and intelligently adjust strategies when encountering obstacles. It is simply the "AI agent team" in the browser. Real-time status updates of the running process allow you to clearly understand the progress of the task, say goodbye to blindness, etc.

Multiple identities, unlock new postures for web automation

It is appropriate to call Nanobrowser a "AI-interactive version of RPA". It not only automates repetitive web tasks like traditional RPA, but also incorporates the wisdom of AI, giving it the ability to understand natural language instructions, make intelligent decisions and handle complex scenarios . You can ask follow-up questions at any time and communicate context with the AI ​​agent, just like having an intelligent assistant on-call. Moreover, all interaction history will be recorded so that you can review and manage it at any time.

Nanobrowser is still under active development, with many exciting new features already planned on its roadmap , such as extending support for more LLMs , enhancing security , optimizing memory usage , supporting session playback , and developing more professional AI agents.

Project: https://github.com/nanobrowser/nanobrowser