Agent search
An agent needs more than ten blue links: an answer it can cite, page text it can read, and several searches in one round trip. This track is that toolkit.
POST /v1/ai-search returns an answer plus the sources behind it, so your agent can quote instead of guess. Providers include DeepSeek (8 credits), Yuanbao (10), Tavily (48) and Exa (21); live status is on the catalog page.
POST /v1/fetch turns any public page into clean Markdown at 3 credits a page (free on failure). Or add include_content to a search call and get the top pages in the same response, 3 credits each.
POST /v1/batch runs up to 5 searches in one request. Each query is billed by the same rules as a single search — failed and empty queries cost nothing — and one failed query does not block the others.
The official MCP server gives your agent list_services, search and fetch. The SKILL.md in the repository tells an agent when to use each one. Setup takes one paste.
Install page →Example: answer with citations
curl https://api.serphubapi.com/v1/ai-search \
-H "X-API-Key: shapi_..." \
-H "Content-Type: application/json" \
-d '{"q": "What changed in the latest EU AI Act guidance?", "provider": "auto"}'The response carries the answer and its cited sources. Billing follows the same rules as search: cache_first is half price, cache hits, failures and empty results are free.
Example: batch search
curl https://api.serphubapi.com/v1/batch \
-H "X-API-Key: shapi_..." \
-H "Content-Type: application/json" \
-d '{"queries": [
{"q": "EU AI Act guidance", "engine": "google"},
{"q": "EU AI Act guidance", "engine": "baidu"}
]}'Up to 5 queries per request. The response returns each query's results with its own meta, plus total_credits_charged for the whole batch.
Use with AI agents →One key. Every search.
1,000 free credits on signup. No credit card.