Firecrawl /interact endpoint allows you to chat with scraped web pages
Firecrawl launched a new endpoint that lets AI agents interact with a live browser session after scraping — logins, infinite scrolls, and dynamic content included.
Firecrawl launched its /interact endpoint today, letting developers scrape a page and then continue interacting with the live browser session using either natural language prompts or raw Playwright code.
Traditional scraping grabs static content and stops. JavaScript-heavy pages, login-protected dashboards, infinite scroll feeds, and multi-step forms have always required a separate browser automation layer. /interact collapses that into a single workflow.
- How it works: Scrape a page to get a
scrapeId, then send natural language prompts or Playwright code to that session. The session stays alive with full state — cookies, login status, scroll position — until you explicitly stop it. - Natural language actions: The AI agent handles clicking, form filling, scrolling, navigation, and extraction automatically from a plain English prompt.
- Persistent profiles: Log in once and resume the same authenticated session later without re-authenticating.
- Live session URL: Every interact session generates a shareable real-time browser view you can watch or manually control.
- Playwright fallback: Drop in raw code on the same session when you need precision beyond what natural language handles.
Sessions are billed separately from scrapes. Stopping the session with a DELETE call releases resources and finalizes billing.
The feature integrates with Firecrawl’s existing /scrape, /search, /crawl, and /extract endpoints, so scrape-interact-scrape loops work natively.
The Bottom Line: The new endpoint makes Firecrawl considerably more attractive for agentic workflows. Agents can now interact with live web pages directly rather than relying on static snapshots, which could improve response quality and reduce token usage at the same time.
Source: Firecrawl
If you need on-demand GPUs for training, fine-tuning, inference, or running open-source models, give RunPod a try.
- Available hardware: H100, H200, A100, L40S, RTX 4090, RTX 5090, and 30+ more
- Cost: significantly cheaper than AWS or GCP, billed per second, no contracts
- Setup: spins up in under a minute, 30+ regions worldwide
Get the core business tech news delivered straight to your inbox. We track AI, automation, SaaS, and cybersecurity so you don't have to.
Just read what you want, and be done with it.
