An easier hello
One recommended starting point. Example questions wait for you to press Send.
Tell Nabsun what you want done. It can navigate websites, click controls, fill forms, and work across tabs. Review requested actions and approve when you are ready.
Windows x64 and Apple-silicon Mac · About 1.4 GB, including the model
All downloads and installation detailsFree and open source. Start with the included local assistant, or connect a more capable assistant for demanding tasks.
Nabsun selects the category and fills the details. You review before approving submission.
Saturday 12:00Page: Service request
Open form → fill fields → review → approve submission
Explore an example. This illustration does not run an AI task.
Open a website, describe the result you want, and let Nabsun work through the steps.
Install Nabsun, then browse as usual.
Try: “Fill this form with these details. Let me review before submitting.” Press Send to start.
Follow its progress as it navigates, clicks, and types. Review the result, approve requested actions, or use Stop to cancel.
The included local assistant needs no account and can handle simple questions and basic browser tasks. For longer tasks with several steps, connect a more capable assistant such as Codex or Claude. Results depend on the assistant and website.
Codex and Claude use an existing eligible account. Online services receive your questions and any page text you include, and their charges and account limits apply. In 0.1.13, choose an assistant in Settings → Assistant choice; older versions call this section Model.
| Option | What you need | Where it runs | Useful for |
|---|---|---|---|
| Built-in Qwen3 1.7B Included | No account or API key | On your CPU | Simple questions and basic browser tasks |
| Codex / Claude Code | Connect in the app and sign in to an eligible provider account | Through the provider’s CLI and service | Tasks that benefit from a more capable model |
| OpenAI / Anthropic API | Your API key; provider usage charges apply | At the selected cloud provider | Choosing models through your own API account |
| Ollama / custom local model | Your Ollama server, or a compatible local GGUF model and engine | On the configured machine or server | Using a local model you already run |
Answers from the built-in assistant can work offline. Loading websites and using online assistants still require internet access.
Recorded on an earlier release. Version 0.1.13 simplifies the header and first-run guide; the recording shows the older interface.
The YouTube player loads when you click. Open on YouTube ↗
Nabsun 0.1.13 is available. Start with a simpler welcome guide, a calmer assistant header, example questions, and a shutdown stability fix. Choose your platform below to install it.
One recommended starting point. Example questions wait for you to press Send.
A compact header, with the other controls in one tidy options menu.
Full-page Settings, clear categories, and advanced controls when you need them.
A fix for delayed window updates that could cause an error after closing.
Attached or detached assistant layouts, saved 75–200% text zoom, queued follow-ups, and Stop remain available. App updates require downloading and installing the new version. Latest published release notes
Download only from the official GitHub release. Files include the local model and are about 1.4 GB. Download buttons use the latest public release, not draft builds.
Nabsun is a developer preview. Windows builds are unsigned. Mac builds are ad-hoc signed, but have no Apple Developer ID signature or notarization. Your computer may warn you or block the app because it cannot verify the publisher. A warning is not proof that a download is safe: check the source and checksum before deciding to open it.
Prefer no installation? Choose the portable .exe, then open it directly. It can show the same warning.
Smart App Control or a work/school security policy may block unsigned apps without offering an exception. Ask your administrator or wait for a signed release. Keep your protection enabled.
If Open Anyway is unavailable, your Mac may be managed by an administrator. Ask them for help. If macOS reports damage or malware, stop and check the release source and checksum; do not treat that as an ordinary publisher warning. Report the exact message in a GitHub issue.
Open SHA256SUMS.txt from the same release as your file. Find its filename and compare the full checksum with the output below. Replace the example filename with your actual download.
Get-FileHash "$HOME\Downloads\Nabsun-VERSION-x64-setup.exe" -Algorithm SHA256shasum -a 256 "$HOME/Downloads/Nabsun-VERSION-arm64.zip"A matching checksum confirms that the file matches that release; it does not establish publisher identity or guarantee safety. If it does not match, do not run the file.
No Intel Mac or Linux installers are currently offered. All published release assets
Nabsun is a developer preview. Review the assistant’s output and actions, especially before submitting forms or changing account data.
Reading pages is automatic. Clicking, typing, navigation, and page code require approval by default. You can allow individual tools or enable Autopilot for write actions. Autopilot stays enabled until you turn it off; page code has its own permission.
Yes. Nabsun exposes browser tools through MCP, a protocol that lets AI tools connect to other applications. Copy the connection configuration from Settings → Integrations → Connect an external agent. Calls use the same approval settings as the assistant sidebar.
The bundled model runs locally, so its inference does not send your prompts to a cloud model provider. Normal browsing still contacts websites, and update checks contact GitHub. If you choose a cloud assistant, it receives the context sent to it. CLI credentials remain in the provider CLI’s store.
You can edit soul.md with preferences such as your time zone or how you like answers. By default this file is shared only with local models. Settings shows its location and whether it is shared with the selected provider.
Open a GitHub issue with your Nabsun version, operating system, and steps to reproduce the problem. Remove passwords, API keys, and other private information from screenshots or logs.
Requires Node.js and npm. These commands install dependencies and fetch the local engine and model. Packaged downloads above do not require these steps.
git clone https://github.com/naveenalavilli/nabsun.git
cd nabsun
npm ci
npm run fetch:model
npm startIf Electron did not download, run npm run setup. Linux requires manual engine setup. See the development instructions.
Start with the included assistant, or connect an external agent to carry out tasks in your open tabs.
Start with the included assistant, ask a short question, and check the answer. No account needed.
Use Nabsun as a browser MCP server for your editor or external agent, working in existing tabs with shared approval controls.