About
Skyvern builds an open-source AI agent and API platform that helps companies automate browser-based workflows using large language models and computer vision. It serves operations and business teams handling procurement, insurance quotes, invoice downloads, and other portal-based tasks, differentiating itself from brittle scripts through adaptive browser-native automation.
Market
Skyvern competes in AI-powered browser automation and workflow automation, positioning itself as an adaptive alternative to traditional scripts and site-specific robotic process automation. Its differentiation is the combination of LLMs and computer vision, enabling workflows to operate across many websites and adapt to redesigns, alongside cloud browsers, an API, an open-source SDK, and no-code tooling.
Skyvern targets companies with repetitive, browser-based workflows, particularly operations and automation teams working across vendor portals and multiple websites. Its likely buyers are mid-market and enterprise operations, procurement, customer-operations, and engineering leaders seeking to automate manual processes without maintaining brittle, site-specific scripts.
At a Glance
Problem
Businesses still perform large volumes of browser-based work manually: employees log into multiple portals, download invoices, copy and reconcile data, fill forms, and navigate legacy systems. Conventional automation often depends on brittle DOM selectors, XPath, APIs, or custom scripts that break when websites change, creating recurring maintenance costs and limiting automation to predictable environments. Skyvern’s core economic promise is to replace this repetitive labor and ongoing script refactoring with adaptable browser agents. Its clearest use cases are portal-based back-office operations, such as automatically logging into supplier or customer systems and downloading invoices, as well as complex procurement and insurance-quote workflows.
Product / Service
Skyvern is an AI browser-automation platform that combines large language models, computer vision, browser automation, and code execution. Users describe a workflow rather than building every interaction by hand; the agent interprets visual page elements, performs multi-step actions, extracts structured data, and can continue working across sites it has not previously seen. Because it does not rely primarily on predetermined selectors, it is designed to withstand website layout changes and handle edge cases such as CAPTCHAs and two-factor authentication.
The product is delivered both as open-source software and as a managed cloud service. The core platform is available under AGPL-3.0, while Skyvern Cloud runs the infrastructure and adds parallel execution, proxy networks, and anti-bot and CAPTCHA capabilities. Developers can integrate it through Python and TypeScript SDKs, APIs, webhooks, MCP, or Docker-based self-hosting, giving customers a choice between a managed service and deployment in their own stack.
Market
Skyvern competes in robotic process automation, workflow automation, and the emerging market for AI browser agents or computer-use systems. Its closest alternatives include Browser Use and general-purpose computer-use agents such as Anthropic Computer Use and OpenAI Operator; it also overlaps with established RPA and integration platforms such as Automation Anywhere and Zapier. Skyvern differentiates around form-heavy, repeatable enterprise workflows, combining visual navigation with a workflow layer rather than offering only a general-purpose browsing assistant.
The company appears to have meaningful early adoption rather than being merely pre-product: its site reports more than 500 enterprise users, 10 million workflows run, 22,600-plus GitHub stars, and a 99.9% uptime SLA. The available evidence does not establish verified revenue, profitability, or a definitive ARR figure; however, the reported user, workflow, and open-source-community metrics indicate substantial product traction. YC identifies Skyvern as an active Summer 2023 company in AI, RPA, workflow automation, and open source.
Founders & Leadership
Funding History
Y Combinator
Recent News
Skyvern compares its browser-automation approach with Firecrawl, emphasizing differences between interacting with websites and reading public web pages.
Skyvern discusses the emerging agentic process-automation stack and the infrastructure needed to operate browser agents at scale.
Skyvern publishes an overview of Browser MCP, explaining how it works and surveying available browser MCP servers.
Skyvern renamed Workflows to Agents across its product, public API, and documentation. The API moved under /agents, including the POST /v1/run/agents endpoint and an agent_id alias.
Make documents its Skyvern integration, including automation for retrieving and processing satellite images for use cases such as agricultural monitoring.
Skyvern announced a $2.7 million seed round to address manual browser work and advance its open-source browser-automation platform and agent architecture.
Skyvern published a guide to AI web agents and browser automation, highlighting its reported 85.8% score on the WebVoyager benchmark and cross-site operation.
Skyvern describes combining AI automation with proxy-network integration for workflows requiring specific locations and adaptation to changing websites.
Skyvern explains its AI scraper workflow and reports that its system can write and maintain its own code, making scraper building 2.7x cheaper and 2.3x faster.
Skyvern announced a verified node available through n8n Cloud, making it easier to connect Skyvern browser automations with n8n workflows.
Active Roles
3Business Model
Skyvern combines an open-source offering with hosted paid plans: free usage, monthly Hobby and Pro subscriptions priced at $29 and $149, and Enterprise plans. Its revenue model is therefore primarily subscription- and usage-based SaaS, supplemented by enterprise contracts.