✍️Writing & Content58🎨Image Generation71🎬Video & Animation120🎵Audio & Music100💬Chatbots & Assistants109💻Coding & Development444📈Marketing & SEO197⚡Productivity401🎯Design & UI/UX120📊Data & Analytics126📚Education & Research55💼Business & Finance174🏥Healthcare & Wellness22🔍Search & Knowledge20🤖AI Agent Infrastructure208🛡️AI Security & Testing32🧊3D & Spatial22🔎SEO Tools113🏡Real Estate7🗃️Data Extraction104🧠ADHD & Focus Tools11🔬Research & Academia45🧩LLM APIs & Models34⚙️Automation & Workflows45🔐Security & Privacy31📊Analytics & BI55⚖️Legal & Contracts14
Listed in AI Agent Infrastructure with 276 other toolsPart of 4557+ curated AI tools on AISO
Alumnium logo

Alumnium

An open-source MCP browser server exposing do(), get() and check() instead of raw primitives — 98.5% on WebVoyager with Claude Code, benchmark published

open-sourceFree and open source; the project publishes its full WebVoyager benchmark results, session transcripts, screenshots, evaluator responses and a reproduction fork rather than only a score. No paid tier is published on the site. Distribution is via GitHub with community support on Discord and Slack; model costs are billed by whichever provider drives the agent.View full pricing →

About Alumnium

Alumnium is an open-source MCP server that gives a general-purpose coding agent high-level browsing rather than raw browser primitives, and it has the benchmark result to argue the design is right. Used with Claude Code and Selenium it scores 98.5% on WebVoyager, above the previous published state of the art of 97.1%, and because the project is open source it publishes the full evidence — session transcripts, screenshots and evaluator responses at webvoyager.alumnium.ai — plus a fork of the benchmark with the code to reproduce it. The architectural argument is the useful part for anyone choosing a browser tool. One approach is a fully autonomous browser agent that takes a goal and returns an answer as a black box, separate from whatever else your agent is doing. The other is exposing raw click, type and navigate primitives, which gives the agent full control but floods the context window with accessibility trees and screenshots on every state change, derailing the main agent from its actual job and triggering compaction. Alumnium sits between the two: it exposes a small set of high-level tools — do(), get() and check() — that compress the knowledge of how to browse into three calls, so the main agent keeps its context for the task it was actually given. It works with Selenium and integrates as an MCP server.

Does ChatGPT recommend your AI tool?

If you're building in AI Agent Infrastructure, run a free AI-visibility scan on your own product — we ask ChatGPT across 5 prompt angles and score how often you get named. ~30 seconds, no signup, no card.

Key Features

✓Three high-level MCP tools — do(), get(), check() — instead of raw browser primitives
✓Keeps accessibility trees and screenshots out of the main agent's context window
✓98.5% on WebVoyager with Claude Code and Selenium, above the prior 97.1% SOTA
✓Full benchmark evidence and a reproduction fork published openly
✓Works as an MCP server alongside Selenium

Tags

mcpbrowser-automationopen-sourceseleniumclaude-code
🏷️

Is Alumnium your tool?

This is the page buyers and AI assistants read when they look up Alumnium. Claim your listing for $19 one-time — no subscription, nothing to cancel — and take a capped slot in your category: each one sells a fixed number, and yours ranks above every free tool in it, with a Featured badge. Prefer it ongoing? Monthly is one click away on the next page.

Stay updated on AI Agent Infrastructure tools — join our weekly newsletter

One concise email with fresh launches, trending picks, and featured standouts.

Alternatives to Alumnium

View all Alumnium alternatives →

More AI Agent Infrastructure tools

Agent connectivity: not yet verified