# Nanobrowser

> Updated 2026-10-02 · type: tool · category: mcp-agents · status: active · rev 1

Nanobrowser is an open-source Chrome extension running multi-agent web automation in your browser with your own LLM key, an alternative to OpenAI Operator.

- Open source: yes (Apache-2.0)
- Self-hostable: yes
- Pricing model: free
- Best for: A growth or ops person who wants to automate real browser tasks — navigating sites, filling forms, pulling data — with an agent that runs locally in Chrome and uses their own LLM key, as an open alternative to a hosted operator agent.
- Not for: A team that wants a fully-managed, hosted web agent with a support contract and no local setup, or one that needs guaranteed reliability on complex sites, where any browser-automation agent will still be brittle.
- Last verified: 2026-10-02

- **Canonical:** https://gtmstacker.com/registry/tool/nanobrowser/
- **Source:** [nanobrowser · GitHub](https://github.com/nanobrowser/nanobrowser)
- **Tags:** mcp-agents, data-scraping, web-automation, browser-agent, byo-llm, self-hostable
- **Repository:** https://github.com/nanobrowser/nanobrowser

## Is Nanobrowser open source?

Yes, Nanobrowser is open source under the Apache-2.0 license.

## How much does Nanobrowser cost?

Nanobrowser is free to use.

## Can I self-host Nanobrowser?

Yes, Nanobrowser can be self-hosted (the source is available under the Apache-2.0 license).

## Alternatives & related

- [BrowserSkill](https://gtmstacker.com/registry/tool/browserskill/)


---

Nanobrowser is an open-source Chrome extension for AI-powered web automation: it runs multi-agent workflows inside your own browser, driven by your own LLM API key, and positions itself as an alternative to OpenAI Operator. Open source: yes (Apache-2.0); self-hostable in the sense that it runs locally in Chrome with your key. It is free OSS, has ~13.9k stars, and is actively developed (~376 commits).

## What it does

Nanobrowser puts a web-automation agent directly in your browser as an extension. You give it a task, and it drives the browser to carry it out: navigating, clicking, filling forms, and pulling information off pages. It uses a multi-agent approach rather than a single loop, and it runs on your own LLM API key instead of a hosted model, so the automation happens on your machine and the model calls go through your account. That is the contrast it draws with OpenAI Operator: the same kind of browser-operating agent, but open source, local, and bring-your-own-model. Open source: yes (Apache-2.0); free OSS; you supply the LLM key.

## Provenance

- Apache-2.0 per repo (license read from GitHub metadata, not guessed); ~13,857 stars; ~376 commits; active (pushed 2026-10-02) (github.com/nanobrowser/nanobrowser, verified 2026-10-02).
- Repo description: "Open-Source Chrome extension for AI-powered web automation. Run multi-agent workflows using your own LLM API key. Alternative to OpenAI Operator."
- Surfaced in the GTM Stacker X/Twitter signal report (2026-10-02 pass); license/facts independently verified 2026-10-02.
- The multi-agent design and the Operator comparison are the project's own positioning (vendor-claim); not independently tested here.

## Why it matters for a GTM stack

A lot of GTM grunt work is browser work: pulling a list off a directory, checking a prospect's site, filling a portal, gathering data a clean API would not give you. Nanobrowser is an open, local way to point an agent at exactly that, on your own model budget, without routing your targets and sessions through a hosted operator service. For a growth or ops person who wants that leverage and wants it owned, it is worth a trial. The honest read: browser automation is brittle by nature, so any workflow you build will need upkeep as sites change, and you carry the LLM API costs yourself. The Operator comparison is positioning, not a guarantee of parity, so test it on your actual tasks before you lean on it.
