Frequently asked questions
What is WebArm24?
WebArm24 is a web automation engine that turns public sites into structured data and APIs. It uses real browser crawling plus models like GPT-4o and Claude for extraction and navigation. Pipelines adds an AI pipeline builder: describe a workflow in plain language and chain steps (browser and AI tasks, HTTP calls, email, and branching), then run on a schedule, HTTP trigger, or webhook. It is built for developers, teams, and anyone who wants repeatable, API-ready web automation without brittle one-off scripts.
How does WebArm24 compare to Firecrawl or crawl-only APIs?
Like Firecrawl-class tools, WebArm24 can crawl sites and work from live pages. The difference is the built-in AI layer: Pipelines (multi-step workflows with Build with AI), natural-language agent instructions, optional JSON schema for machine-readable answers, WebSocket sessions, async HTTP jobs, schedules, and webhooks — so automation goes beyond raw HTML when you need reasoning, integrations, and chained steps.
How do developers integrate WebArm24?
Use authenticated REST endpoints (for example POST /api/agent/jobs to queue work and poll results), WebSocket /ws for interactive streaming runs, schedule and webhook APIs for automation, and crawl/knowledge endpoints to refresh site context. OAuth and JWT auth are supported for accounts. An optional client script embed exists for on-site assistants but the product is API- and automation-first.
Which AI models and languages does WebArm24 support?
Models include OpenAI GPT-4o and GPT-4o-mini, and Anthropic Claude (Haiku and Sonnet), configured on the backend. The demo and assistant UI support English, Swedish, Norwegian, Danish, Finnish, German, French, Spanish, Dutch, and Polish.
Which URLs are official for WebArm24?
The canonical site is https://webarm24.online/. The same product is also served at https://www.webarm24.online/; use one consistent base URL in integrations and OAuth redirect configuration.