Everywhere
AI Assistant that flows with your desktop.
223 followers
AI Assistant that flows with your desktop.
223 followers
Everywhere is dedicated to liberating AI from browser tabs and standalone apps, making it a ubiquitous, native capability of your operating system. We believe true productivity gains stem from the seamless integration of AI with your current tasks.
Launch Team / Built With
Floot MCPBuild and ship web and mobile apps inside Claude or ChatGPT
Promoted
@artemis_li For folks deep in Obsidian or Notion, how seamlessly does this integrate with their internal links?
@masump It reads what’s on your screen in Obsidian or Notion just fine, works well for summaries or quick help. It doesn’t hook into internal links though. That said, MCP tool support is on our roadmap, so deeper integrations will come soon.
@artemis_li Does Everywhere store screen content? How long? Can you guarantee nothing sensitive leaves my machine?
Screen reading AI that lives in the OS is powerful but also a massive security risk. Especially someone coming from corperate that cant just install something like this on my machine, as cool as i think it is haha.
@berghoffer Everywhere doesn’t upload or log anything and all screen context stays local and gets deleted when you delete the chat. What it can “see” depends on the app (some only the visible area, some none).
You can run it with local models like Ollama or LM Studio, so nothing leaves your device.
It’s designed for personal use right now, so no enterprise-grade permission management yet, but privacy and local control are at the core of how it works.
This one looks seamless and very easy to use. May I know more about the scope, your target audience - For whom was this built and what are the pain points you guys were trying to solve
I believe AI needa fit into people's work flow and Everywhere really provides the access to it! I will try it out to see how it works!
I can see the true value if the AI can be more proactive and not reactive, having context of what I can say on screen is a strong prerequisite to achieve it. Like this product!
Added support for MCP (Model Context Protocol)!
You can now add your own MCP tools, which will be automatically enabled when called by an assistant. Supports Stdio, Streamable HTTP, and SSE protocols.
Have a try in the latest version.
@sarrah Under active development. Mac support is our highest priority now.
Congrats on the launch! This looks like a really innovative way to integrate AI seamlessly.
How does Everywhere handle privacy and data security when operating across different applications?
@sanskarix
Everywhere is open-source, meaning all code undergoes community review and security vulnerabilities can be addressed promptly.
Everywhere supports configurable LLMs, allowing you to use local model providers like Ollama or LM Studio.
Our upcoming memory feature will prioritize local embedded models and databases, enabling you to control all data storage and processing locally.
@dearva I liked the third point. All the best!