Start with the result you need
Ask Noema to compare information across permitted pages, follow a useful link or prepare a draft. The task starts in your normal conversation. Progress, questions, action approvals and the result stay together in chat.
Keep control of the task
Choose which sites Noema can access and set a session spending limit. Review actions as the task progresses, stop when you need to, or disconnect to end access. Browser tasks use cloud AI and your Noema credits, without separate model-provider accounts or keys.
Source you can inspect
The Noema Browser extension source is now public under GPL-3.0-or-later. The versioned download includes editable source, licenses and checks, with a checksum alongside it. The repository covers the extension; the Noema application and model service are separately provided.
Install it on your desktop
Download version 0.1.0 from the GitHub release, extract it, enable Developer mode on your browser’s Extensions page and choose Load unpacked. Select the extracted folder containing manifest.json, permit a site and connect through Browser tasks in Noema. This release supports desktop Chromium browsers with side-panel support and uses manual updates. It is not a Chrome Web Store listing.
One more part of the workspace
The extension is optional. Ordinary chat, local AI and the rest of the Noema workspace remain usable without it on desktop and mobile. Browser automation itself is not available in Safari, Firefox or mobile browsers. Broader agents and the future confidential Network remain separate roadmap milestones.
