Repository navigation
skyvern alternatives
Skyvern keeps one of the most honest sentences in this category near the top of its README, and this page is organized around it: "All of the core logic powering Skyvern is available in this open source repository licensed under the AGPL-3.0 License, with the exception of anti-bot measures available in our managed cloud offering." Read on 2026-09-03, at roughly 22.9k GitHub stars. Most projects bury a trade like that in a pricing page; Skyvern states it in plain text, and it deserves credit for doing so. It also means the search that brought you here is usually not "is Skyvern good" - it is - but "which half of that sentence is a problem for me, and what answers it".
Disclosure before anything else: this wiki belongs to invisible_playwright_mcp, an open-source agent that competes with Skyvern and appears below as one of the alternatives. Every claim about Skyvern here traces to its own repository, and where Skyvern is the better tool this page says so.
Python, Playwright underneath, AGPL-3.0, installable via pip or Docker Compose, with Skyvern Cloud as the hosted option. The architectural bet is vision: LLMs read the rendered page and plan actions against what they see, rather than depending on XPath or CSS selectors, and the README describes the result as a "Playwright-compatible SDK that adds AI functionality". On top of that sit parameterized workflows: multi-step automations you define once and rerun with different inputs. The bet - pay tokens to look at pixels so that a site redesign does not break your script - is coherent, and for form-heavy pages that change layout often it is the right one. Nothing below disputes any of this.
AGPL is strong copyleft that extends to network use: if you build Skyvern into a service you offer to other people, you take on source-sharing obligations that MIT and Apache tools do not carry. That is a one-line summary, not legal advice, and for an individual automating their own browser work the license costs nothing at all. But it is the single most common reason companies evaluating Skyvern end up on a page like this one, and it is why every other tool named below is MIT or Apache-2.0.
The part of the sentence after the comma is the part self-hosters feel. What
the README calls anti-bot measures is precisely the piece kept out of the
repository, and two fair things need saying about that. First, it is a
legitimate design: code of that kind is kept private by essentially everyone
who ships it, because publishing it shortens its useful life, and the cloud
is what funds the open core. Second, the consequence for you: what pip install skyvern runs locally is standard Playwright Chromium, and its
sessions answer a page's questions the way any stock automation build does.
That is not a Skyvern defect - it is the shared box of the whole
Playwright-Chromium class, documented on this wiki for browser-use in
what configuration can and cannot change.
And before you attribute a challenged run to any tool, read
why an agent gets blocked: machine
and network facts decide most of these outcomes, and nothing on this page
changes your IP.
Roughly 112.2k stars, MIT, Python. Not a workflow builder: you construct an
Agent with a task and a model, call run(), and its loop does the rest -
a conversational agent rather than a parameterized pipeline. If the license
half was your problem, this is the most direct swap in the same
Playwright-Chromium class. If the cloud half was, know that the structure
repeats: browser-use's README points production users at Browser Use Cloud
and describes it with features including what it calls built-in stealth and
proxy rotation - vendor claims we did not test. The MIT core runs locally
either way. Full treatment on
browser-use alternatives.
Roughly 24.1k stars, MIT, TypeScript with Python and Go SDKs, built by
Browserbase. Three primitives - act(), observe(), extract() - on a
Playwright-style API, under the banner "Playwright was built for testing,
Stagehand is built for agents". This is the alternative for people who liked
Skyvern's Playwright compatibility more than they liked adopting a platform:
you keep your own loop and mix AI steps with plain deterministic calls in
the same script. The full referee comparison is
Stagehand vs browser-use, and the hosted
side of that ecosystem is covered in
Browserbase alternatives.
UI-TARS (roughly 11.4k stars, Apache-2.0) is not a framework but a vision-language agent model: the GUI-operating capability is trained into the weights, and you bring your own scaffold or use its desktop companion. It is the furthest departure on this page - no workflow engine, real inference costs, research-grade edges - but if your read on Skyvern was "vision-driven operation is the right idea", this is that idea without a platform around it. Context in open-source computer-use agents.
Roughly 30k stars, MIT, and the conflict of interest from the top of the page applies to this section. The contrast maps directly onto Skyvern's sentence: where Skyvern's README places its anti-bot measures in the managed cloud, invisible_playwright_mcp's equivalent hard part is the browser itself - a Firefox patched at the C++ level (the invisible_playwright engine) that presents a normal desktop fingerprint - and it ships inside the open-source product, running on your machine. That is a different placement of the same problem, not magic: we make no promise of non-detection anywhere, sites remain free to challenge any visitor, and the engine wiki documents how to test the difference yourself rather than asking you to take a claim on faith. The interaction model is also different in kind: invisible_playwright_mcp is a conversational agent that plugs into an MCP assistant you already run (Claude Code, Claude Desktop, Cursor) or its own local UI, not a vision-workflow engine, and it runs on Windows and Linux only.
Fairness requires its own section. The vision-workflow model is the most resilient approach on this page to layout churn, and none of the alternatives replaces it like for like: browser-use gives you an agent, not rerunnable parameterized workflows; Stagehand gives you primitives, not a workflow engine; invisible_playwright_mcp gives you a different browser and a conversational loop, not structured pipelines. Skyvern also has a hosted option for teams that want managed scale, a straightforward pip and Docker story, and the rare virtue of a README that tells you its own limits. If neither the license nor the placement of the hard part bothers you, stay on Skyvern and close this tab; it is a good tool.
Is Skyvern open source? The core is, under AGPL-3.0. Its README states the exception itself: the measures it describes as anti-bot are available in the managed cloud offering, not the repository.
Skyvern vs browser-use - what is the actual difference? Same class (LLM agents driving Playwright Chromium), different models of use: parameterized vision-workflows versus a conversational agent loop, AGPL versus MIT. Neither changes what the browser underneath looks like.
Is Skyvern free to self-host? Yes, via pip or Docker, plus whatever your model provider charges in tokens. The cloud is a paid product.
Can self-hosted Skyvern handle sites that challenge automation? Its README reserves that category for the cloud. Self-hosted, you are running standard Playwright Chromium, and outcomes mostly track the machine and network facts covered in why an agent gets blocked.
Is invisible_playwright_mcp better than Skyvern? Not in general, and this is our own wiki saying so. It differs where the browser itself or the AGPL is your problem; Skyvern is better where workflow authoring, vision resilience to layout change, or a hosted option matter.
See also: Stagehand vs browser-use for the SDK half of this page expanded, browser-use alternatives for the mirror-image survey, and Choosing an AI browser agent for the full decision framework.
- The Skyvern repository, retrieved 2026-09-03: star count, license, the quoted README sentence, install paths, and the vision-LLM and Playwright-SDK descriptions.
- The browser-use repository, retrieved 2026-09-03: star count, license, the agent API shape, and the cloud description quoted attributively.
- The Stagehand repository, retrieved 2026-09-03: star count, license, languages, and API primitives.
- The UI-TARS repository, retrieved 2026-09-03: star count, license, and model positioning.
Written while maintaining invisible_playwright_mcp, which competes with Skyvern. That is why the page opens by crediting Skyvern's honesty, closes its survey with where Skyvern wins, and quotes the README instead of paraphrasing it.
- OpenAI Operator alternatives
- Open-source Operator-style agents
- Is OpenAI Operator still available?
- OpenAI Operator vs Claude computer use
- browser-use alternatives
- Choosing an AI browser agent
- Open-source AI browser agents
- Open-source computer-use agents
- What is an AI web agent?
- AI browser agents vs traditional scraping
- Cloud browser infrastructure for AI agents, explained
- Browserbase alternatives
- Firecrawl vs an AI browser agent
- Skyvern alternatives
- Stagehand vs browser-use
- Project Mariner is gone: what replaced it
- Manus alternatives
- Gemini computer use vs Claude computer use
- A stealth browser MCP, reviewed honestly by its own wiki
- AI browser vs AI browser agent: which one do you want?
- AI browser agent vs RPA: which one fits the job
- AI browser agent vs n8n, Zapier and Make
- Vercel agent-browser alternatives, compared honestly
- What is an agentic browser? Definition and the two kinds
- Open-source agentic browsers: the three layers, compared
- Choosing an MCP server for browser automation: four axes
- Stealth MCP servers compared: Camoufox, nodriver, Patchright
- Playwright MCP alternatives, and the three you don't need
- Autonomous browser agents: the four rungs of autonomy
- What is actually free in the AI browser agent stack
- browser-use on GitHub: what the repo actually gives you
- Playwright MCP vs Chrome DevTools MCP: different jobs
- How to choose among MCP servers: a map by category
- Which MCP servers are worth adding to Claude Code
- MCP on GitHub: finding servers and judging them fast
- MCP vs an API: the decision, and what the wrapper costs
- MCP alternatives: when the protocol is the wrong shape
- Why does my AI agent get blocked?
- The timing signal AI agents give off
- Agent retry loops trip rate limits, not fingerprints
- Claude computer use detected as a bot
- browser-use getting blocked: what you can and cannot change
- Playwright MCP session blocked: four causes, four fixes
- Playwright MCP and captchas: what actually gets you past
- Cloudflare and a browser MCP server: what is being read
- Can an AI agent solve a captcha? The honest answer
- Getting an AI agent to fill out forms
- The best model for an MCP browser agent, and what it really costs
- Browser problem or model problem?
- How to use a Playwright MCP server with Claude Code
- Extracting data to a CSV with an AI agent
- Monitoring a page for changes with an AI agent
- How to let Claude Desktop control a browser
- How to add a browser to Cursor as an MCP server
- Using an AI agent to hunt for apartments
- Getting website data into Google Sheets with an AI agent
- Using an AI agent to download invoices from portals
- AI agents for web research
- Using an AI agent to test your own website
- How to add a browser to Cline as an MCP server
- Posting to social media with an AI agent
- Posting to Facebook with an AI agent
- Posting to Instagram with an AI agent
- Posting to X with an AI agent
- Automating LinkedIn posts: read this first
- Appointment bots: what they are and what an agent can legitimately do
- Track prices across sites with an AI agent
- Build a lead list with an AI browser agent
- Run an AI browser agent on a schedule
- AI browser agent with a local LLM: what changes
- Should you log your AI agent into your accounts?
- How to write a task an AI browser agent can follow
- Move data between two web apps with an AI agent
- The MCP server
- How the tools are shaped, and why
- Playwright MCP vs the Playwright CLI: which fits when
- Playwright MCP: browser is already in use, and the fix
- Playwright MCP best practices: four decisions that matter
- Playwright MCP with a proxy, and the three leaks it leaves
- A browser MCP server in GitHub Copilot: setup and limits
- Using a browser MCP server for web scraping: the pattern
- Which LLM for browser automation: the four properties
- How to build a browser agent, and what to take instead
- Getting an AI agent to log into a website: three routes
- MCP tools, resources and prompts: who controls each
- How many MCP tools is too many? The context arithmetic
- How to build an MCP server: the decisions, not the scaffold
- Local or remote MCP server: what changes, and what does not
- Writing an MCP client in Python: the thirty-line version
- Self-hosted AI agent: what one actually costs to run
- How long an AI browser agent takes per step, measured
- Text, HTML, snapshot or screenshot: what the agent should read
- Giving an AI browser agent a stopping condition
- Keeping an AI browser agent out of destructive actions
- Why did the AI agent click the wrong thing
- When the page changes under the AI agent
- Running one AI agent task across a list of sites
- Seeing a page as it appears in another country
- Getting data out of a dashboard with no export button
- Two browsers in one session: main and support
- Finding the dead links on a site with an AI agent
- Filling a CRM record from a company's website
- One form submission per spreadsheet row, with an AI agent
- Dated screenshots of a page as evidence
- Checking order and delivery status with an AI agent
- Reading a PDF that opens inside the browser
- Summarising a long page or thread with an AI agent
- Collecting every image on a page with its caption
- Collecting event and course listings with an AI agent
- Cancelling a subscription with an AI agent
- What an AI agent can and cannot do inside an iframe
- Shadow DOM and an AI agent: you can click it, you cannot read it
- How to add a browser to Codex as an MCP server
- What a page snapshot costs, per control
- How to let Gemini CLI use a browser
- Native selects and the ones that only look like selects
- Clicking by selector or by coordinates
- How long the agent waits before it gives up
- What a second browser costs
- Uploading a file with an AI agent, and why this one cannot
- Watching the agent work, and when it is worth it
- When not to use an AI browser agent
- Agent or script: deciding once instead of every time
- Using the keyboard instead of the mouse
- Secrets in an agent task: where they end up
- What an agent run should log
- Deduplicating what an AI agent collects
- Normalising values across sites
- Validating an AI agent's output
- Reading a table with an AI agent
- Driving a site's own search and filters
- The task works headed and fails headless
- How to let the AdaL CLI use a browser