Scalebrowser

About Scalebrowser

Scalebrowser is self-hosted browser infrastructure built for AI agents. It runs a patched Chromium fork on the user's own hardware, giving each agent an isolated profile with a persistent identity. The tool keeps cookies, credentials, passkeys, and login sessions intact across runs, and handles email/SMS codes, TOTP, captchas, and browser prompts without human intervention.

Review

Browser agents often fail before they reach the useful part of a task - they get detected at the connection layer, lose their session state, or stall on verification screens. Scalebrowser addresses these failure points as a unified system rather than a collection of patches on top of Playwright. It's self-hosted infrastructure that maintains persistent browser profiles and handles the authentication hurdles that typically break automated workflows.

Key Features

• Persistent, isolated browser profiles that retain cookies, credentials, passkeys, and task memory between sessions
• Built-in handling for email codes, SMS codes, TOTP, captchas, and browser permission prompts without handing control back to a person
• Agent connectivity over MCP, REST, Node.js, Python, or direct CDP with a compact page map that transmits only what changed
• Step-level recording with screenshots, plus automatic secret filtering before page content reaches the model
• Completed runs can render as product demo videos with camera movement, cursor motion, voiceover, subtitles, and channel-ready exports - or as unbroken proof trails

Pricing and Value

Pricing starts at €19 per month, with a seven-day trial available. Every plan includes all features and unlimited profiles. Scalebrowser is self-hosted - browser data stays on the user's own machines - but it is not open source or offline. A paid account controls licensing. Cloud hosting is on the roadmap, but not yet available. GDPR compliance is also planned, though the builder has stated they are still determining how the EU AI Act will apply.

Pros

• Persistent profiles mean agents don't reappear as different devices or lose login state mid-task
• Runs on your own hardware, so browser data never leaves your infrastructure
• Multiple connection methods (MCP, REST, CDP, Node.js, Python) fit into existing agent stacks
• Recording and video export features can double as documentation or client deliverables
• All features are included on every plan - no feature gating by tier

Cons

• Not open source; the browser fork is patched but the licensing model requires a paid account
• Cloud hosting is not yet available, so users must provision and manage their own hardware
• Not well suited for teams that need a fully managed, off-the-shelf cloud service today - self-hosting adds operational overhead

Scalebrowser fits teams and solo developers who already run browser agents and keep hitting detection, login, or session persistence walls. It's most practical for those comfortable with self-hosted infrastructure who want their agents to maintain stable identities across long-running tasks. The video export feature also makes it usable for teams that need to produce walkthrough recordings or audit trails from the same runs.