Browserbase

Provides cloud browser automation capabilities using Stagehand and Browserbase, enabling LLMs to interact with web pages, take screenshots, and run parallel sessions in a cloud browser environment.

Apify

Web scraping and automation platform. Run scrapers, extract structured data, and automate browser-based workflows.

WebRun gives AI agents a real Chrome browser running in the cloud. Instead of wiring up an API for every service, your agent does the work the way a person would: it opens the site, signs in, clicks through the flow, and reports back. Sessions run on persistent environments that keep their Chrome profile, cookies, files, and desktop state, so an agent stays logged in across runs instead of re-authenticating every time. Every session streams live, and a human can take control mid-task to clear a CAPTCHA or enter a credential, then hand it back.

Give your AI a real computer and let it finish the job. Noxgild gives agents secure access to authorized Windows, macOS, and Linux computers plus a hosted Cloud Browser, so they can actually complete work end to end: write and modify files, run commands, manage long-running processes, use signed-in websites, operate native apps, test software, investigate issues, deploy changes, and execute multi-step workflows. Keep using the AI tools you already know while Noxgild provides the execution layer—with approvals, credentials, and protected human gates remaining under your control.

Hosted Playwright browser automation for AI agents. Returns accessibility trees instead of screenshots, cutting token usage by 77%. Navigate, interact, extract structured data, and take screenshots — all via MCP. Zero infrastructure, credit-based pricing.

Reduck MCP allows agents to discover, run and create browser automation scripts that serve as tools. Scripts run in your own Chrome, through our extension, which allows your agent to work where you are logged in — no credentials exposure, and same fingerprints so no bot detection.

Enable AI assistants and applications to perform automated web browser interactions, testing, and scraping through a standardized interface. Control multiple browsers with comprehensive element interaction and advanced navigation tools. Benefit from type-safe, robust browser automation powered by Selenium WebDriver.

Scrape any page or document to markdown, map and crawl whole sites, and drive a real browser — all from one server. Pay per call with a wallet. No API key, no signup, no monthly plan. Keyless clients get a free daily allowance so the first call works before any account exists. What it does that a scraper usually does not A map of the controls — every link, button and field with its role, its readable name and where it sits on screen. The map an agent needs to decide what to click. Browser sessions that stay open — look, click, look again, without reloading and without losing state. Documents by url — PDF, Word, Excel and CSV come back as markdown, no browser involved. Every output in one call — markdown, html, links, controls, accessibility tree, elements by selector, screenshot, PDF, or AI-extracted JSON. Tools: webscrape · webmap · webcrawl · websessionopen · webact · webscrapebatch · websearchexa · and status, c

Give your AI agent the live web. One MCP server for Search, Fetch, Browser, and Agent: browse real sites, extract structured data, fill forms, and run multi-step workflows. Search and Fetch are free.

Kernel

Deploy and manage Kernel apps, automate browsers, and search documentation with Kernel. Securely run workflows, monitor deployments, and spin up sandboxed browser sessions, all instantly accessible through a hosted service built for AI-native automation.

Scrayle.com gives AI agents full web access through 35 tools spanning scraping, browser automation, AI-powered browsing, workflow orchestration, and more. • Scrape & Crawl - Extract content from single pages or crawl entire sites. Batch scrape multiple URLs at once. Get clean markdown, HTML, links, and metadata. • Cloud Browsers - Launch stealth cloud browsers with anti-detection. Navigate, click, type, screenshot, scroll, select dropdowns, and execute JavaScript - all remotely. • AI Browser Agent - Let an AI agent browse the web autonomously. Give it a task and it figures out the clicks, scrolls, and form fills on its own. • Workflows Create, run, and manage multi-step automation workflows. Chain scraping, browser actions, and data processing together. • CAPTCHA Solving Solve reCAPTCHA, hCaptcha, Turnstile, and other challenges automatically. • Marketplace - Browse and run pre-built scraping templates from the Scrayle marketplace. • Key-Value Storage - Store and retrieve data

Headless browser via Playwright. Scrapes JS-rendered pages, takes screenshots, extracts links and tables, fills forms, runs arbitrary JS. Useful when fetch() is not enough.

Any website as MCP tools: Unbrowse replays a site's own first-party API server-side and returns verified results — no browser on your side. Search thousands of indexed sites, run their tools, read pages, sign in from your saved logins, and learn new sites in a recorded cloud browser.

Zero-configuration on-demand cloud browser for your agent. Navigate web pages, fill forms, and capture screenshots effortlessly with pay-as-you-go pricing at just $0.10 per hour. Exposes Chrome Devtools MCP interface.

Agent-first browser server: one Rust binary with V8 built in. Tiered fetch, multi-engine search, screenshots, downloads, click/eval, persistent login sessions, and render_markdown (markdown + archify diagram fences to deterministic HTML+SVG artifacts). Stealth TLS fingerprints, optional 2captcha integration, no Chromium.

Turns any webpage into a structured action manifest — a list of what's clickable, fillable, and submittable, not just what's on the page. Built for agents that need to act on a page, not just read it.

Use your Mac, Windows (beta) or Linux computer from ChatGPT, Claude and other AI apps: read and organize files, edit Excel and PDFs, run commands, and on a Mac edit videos and build iOS apps. Computers start read-only; you turn on each permission per computer.

Fetch any public web page through managed proxies, with optional JavaScript rendering, wait conditions and CSS or AI extraction rules, as markdown, text, HTML or JSON. Runs on HasData's hosted API.

Glazyr Viz bypasses fragile DOM scraping and Cloudflare blocks by feeding raw Chromium memory frames directly to your AI agent. Achieve 177 FPS zero-copy vision with built-in, agent-native USDC compute settlement. The Screenshot Tax is over.

Web access for AI agents: search, scrape, crawl, automate browsers, and extract AI responses.

Give your LLM a browser through MCP. Navigate, read pages, use visible controls and inspect AI-built website evidence in your own isolated Kitsuvo browser. Install Kitsuvo for Mac or Windows, run the local cloud connector, and explicitly approve OAuth access. The relay works only while your connector is running. Private and sign-in pages are refused. Client-requested page data is shared with the connected assistant. Setup: https://kitsuvo.com/integrations