Page source mode needs browser mode enabled but disables JS rendering, so JS-built pages come back empty. Use the normal browser mode output for those, and base64 encode js_snippet when calling the raw API.

Context: Official docs (ScrapingAnt request/response format): the page source mode requires browser mode to be on, but it turns off JS rendering. What you get back is the DOM without JS execution, so pages that build content client side come back empty. If you need the rendered content, use the normal browser mode output instead. Also note: if you use the raw API, the js_snippet parameter expects a base64 encoded snippet, not raw JS.