---
title: "Scalebrowser Blog, page 5"
description: "Guides and measurements for teams whose AI agents drive a real browser: detection, captchas, sign-ins and infrastructure."
canonical: "https://scalebrowser.net/blog/page/5"
---

> ## Documentation Index
> Fetch the complete documentation index at: https://scalebrowser.net/llms.txt
> Use this file to discover all available pages before exploring further.

# Scalebrowser Blog, page 5

> Guides and measurements for teams whose AI agents drive a real browser: detection, captchas, sign-ins and infrastructure.

- [Automate 2FA for an AI agent: 6 digits every 30 seconds](https://scalebrowser.net/blog/automate-2fa-totp.md): How an AI agent passes two-factor authentication: the 6-digit TOTP code computed per RFC 6238 inside its profile, and why cloud guides say to disable 2FA.
- [Accessibility tree for LLM agents: 462 tokens per step](https://scalebrowser.net/blog/accessibility-tree.md): What the accessibility tree is, why LLM agents read it instead of HTML, what a page map built from it costs per step, and how it catches a page's error message.
- [Cloud browser or local browser for AI agents?](https://scalebrowser.net/blog/cloud-browser.md): What a cloud browser for AI agents costs at 5 vendors: session caps, hour and gigabyte meters, GPU and fingerprint, set against a browser on your own machine.
- [Browser agents: how sites detect them and how to pass](https://scalebrowser.net/blog/ai-agent-browser.md): What a browser agent is, the 4 layers a site reads to detect the browser it drives, which stacks exist today, and what stops an agent after detection.
- [AI agent login: the 4 walls before a signed-in session](https://scalebrowser.net/blog/ai-agent-login.md): How an AI agent signs in to websites and stays signed in: the second factor, the emailed code, the passkey and the lost session, each with what we measured.
- [Behavioral biometrics: 11 mouse measures against 120 people](https://scalebrowser.net/blog/behavioral-biometrics.md): How behavioral biometrics tell an automated mouse and keyboard from a person, the 11 pointer measures we track against 120 people, and the 5 still outside.
- [Bot detection: 4 layers that flag an agent's browser](https://scalebrowser.net/blog/bot-detection.md): How bot detection reads a browser on 4 layers before and after the page loads, how the layers add up to one verdict, and where our own runs went wrong.
- [Browser fingerprint test sites: what 8 detectors check](https://scalebrowser.net/blog/browser-fingerprint-test.md): 8 browser fingerprint test sites and what each one reads, how long each takes, and the results that flag even a real Chrome if the test is set up wrong.
- [What browser fingerprinting reads](https://scalebrowser.net/blog/browser-fingerprinting.md): What browser fingerprinting reads, which values have to agree with each other, why a random value stands out, and how to check your own browser.
- [Browser MCP servers compared: tokens per step across 6 tools](https://scalebrowser.net/blog/browser-mcp.md): Six browser MCP servers measured on live pages: tokens to see a page, tokens per step, the tool list's fixed cost, and which server fits which job.
- [A Browserbase alternative on your own machine](https://scalebrowser.net/blog/browserbase-alternative.md): Browserbase compared with a browser on your own machine for AI agents: session length, billing, proxies, fingerprint, second factor and passkeys, as documented.
- [Camoufox review: Firefox patches and a new device per launch](https://scalebrowser.net/blog/camoufox.md): What Camoufox's C++ Firefox patches do well, why a new fingerprint at every launch hurts a returning login, and the 2025 build 63.4% of downloads still run.

Page 5 of 7. Previous: [page 4](https://scalebrowser.net/blog/page/4.md). Next: [page 6](https://scalebrowser.net/blog/page/6.md).