ScrapingBee is a scraping company. The Trends path is "send this trends.google.com URL, get HTML or extracted JSON." Google properties need custom_google=true. Proxies, rendering, and CAPTCHA solving are the product. Trends API never takes a URL. POST https://api.trendsapi.ai/api takes mode and source. body is a JSON string and must be parsed a second time. Docs: the API reference. Caps: pricing.
custom_google on a Trends URL
A request without that flag can be refused for Google hosts. That flag is why a Trends fetch is not priced like a static page. ScrapingBee's own Python Trends helper documented 15 credits on a trending-now style call when checked by third parties. Treat credit cost as something to measure on the current plan, not as a constant.
Credits per rendered page
GET https://app.scrapingbee.com/api/v1
?api_key=KEY
&url=https://trends.google.com/trends/explore?q=coffee
&custom_google=true
That is a page. render_js, stealth_proxy, and country_code are scrape knobs. They do not appear on Trends API.
POST https://api.trendsapi.ai/api
Authorization: Bearer <api_key>
Content-Type: application/json
{"mode":"get_time_series","source":"google search","keyword":"coffee"}
extract_rules versus a mode field
ScrapingBee can return raw HTML or apply CSS / AI extract rules. The schema is whatever the page contains this week. get_growth returns named windows. get_top_trends returns a ranked board. Those schemas do not track a CSS class rename.
HTML fidelity versus a 0-100 series
Keep ScrapingBee when the team already buys credits for Search and News and must screenshot or extract a widget this API does not expose. Move when the warehouse key is source plus keyword and the next join is TikTok. Do not send a Trends API Bearer key to app.scrapingbee.com.