Skip to content

AI Agent Infrastructure · Zero-Signup x402 Base USDC ($0.001) · Remote MCP OAuth 2.1 · SKILL.md

Real-Time Google SERP, News & Trends Built for AI Agents and RAG

Give autonomous research loops, Claude Code, Cursor, LangGraph, and OpenAI Agents deterministic live Google Search rankings, People Also Ask Q&As, breaking publisher news, 54-week trend curves, and <80ms query expansion at $0.08–$0.20 / 1,000 calls — 40x cheaper than Tavily and 75x cheaper than SerpApi. Supports zero-signup x402 USDC pay-per-call ($0.001 SERP/News/Trends, $0.0005 Autocomplete) on Base Mainnet (eip155:8453) or free API key.

NEW · Native x402 Zero-Signup Micropayments (Base Mainnet USDC · eip155:8453)Coinbase CDP Facilitator · Bazaar Discovery · $0.00 Settled on Error/Empty

Autonomous agent wallets can call https://api.searchdataapi.com/v1/search (and https://api.searchdataapi.com/x402/v1/*) with zero signup and zero API key: pay $0.001 USDC per SERP/News/Trends call or $0.0005 USDC per Autocomplete call on Base Mainnet via HTTP 402 PAYMENT-REQUIRED / PAYMENT-SIGNATURE, or claim 3,000 Free Credits ($0.30, no card) with a standard sr_live_ API key.

OAuth 2.1 · PKCE S256 · Read-Only Scope (mcp:tools:read)3,000 Free Credits ($0.30) Auto-Granted

Claude Web / ChatGPT 1-Click Remote MCP URL

Paste into Claude.ai (Settings → Integrations → Add Custom Connector) or ChatGPT Connectors for zero-key 1-click Google / Email sign-in:

https://mcp.searchdataapi.com/v1/mcp

Using Cursor, Claude Desktop, Claude Code CLI, or Windsurf with an x-api-key header? Connect to https://mcp.searchdataapi.com/mcp or see the full MCP setup guide →

Official Anthropic MCP Registry IdentifiersFull MCP Tool Schema Reference (/mcp) →
com.searchdataapi/google-search-serp-news-trends-autocomplete-mcpcom.searchdataapi/x402-google-search-serp-news-trends-mcpcom.searchdataapi/mcp-server

Zero-Boilerplate Integration

Connect Any Agent Runtime in 30 Seconds

Mint an instant sr_live_... key with 3,000 Free Credits ($0.30, no card) and drop SearchData API into Claude Code, Cursor, LangChain, LlamaIndex, or the OpenAI Agents SDK.

// Zero-signup x402 USDC pay-per-call on Base Mainnet (eip155:8453):
// $0.001 / call (SERP, News, Trends) · $0.0005 / call (Autocomplete) · $0.00 if empty/error
import { wrapFetchWithPayment } from "@x402/fetch";

const fetchWithX402 = wrapFetchWithPayment(fetch, agentBaseMainnetClient);
const res = await fetchWithX402("https://api.searchdataapi.com/v1/search?q=openai&country=us&limit=10");
const data = await res.json(); // Also available via https://api.searchdataapi.com/x402/v1/search?q=openai

// Or inspect raw HTTP 402 + PAYMENT-REQUIRED + Bazaar Discovery schema directly:
// curl -i "https://api.searchdataapi.com/v1/search?q=openai"

Built for Multi-Turn Agent Loops

40x Lower Tool-Call Cost & Deterministic Error Semantics

$0.0002 / Tool Call

40x Cheaper than Tavily · 75x Cheaper than SerpApi

Deep research agents routinely fire 8–15 parallel queries per user question. At $0.20/1K PAYG ($0.08/1K Scale), 10 searches cost $0.002 instead of $0.08–$0.16 on Tavily or $0.15 on SerpApi.

12ms Warm · 1,120ms Live

Synchronous JSON Without Async Task Queues

Zero Docker Actor cold starts and zero two-step task_post polling. Inspect live p50/p95 histograms on /benchmarks.

Agent-Safe Rate Headers

60–600 RPM + Explicit Retry-After

Burst-smoothed token buckets handle parallel agent fan-out cleanly, emit standard Retry-After headers on HTTP 429, and charge 0 Credits on empty results.

3 Copy-Paste Production Blueprints

End-to-End Autonomous Agent Workflows

Workflow 01 · GET /v1/search + GET /v1/news/search

Hallucination-Free Deep Research & Fact-Checking Agent

Concurrently query Google Organic SERP + People Also Ask alongside the live Google News wire to ground LLM answers with canonical URLs and UTC-timestamped publisher citations ($0.0004 total per dual-rail run).

import asyncio
import os
import httpx

API_KEY = os.environ["SEARCHDATAAPI_KEY"]  # 3,000 Free Credits ($0.30) at /signup
HEADERS = {"Authorization": f"Bearer {API_KEY}"}

async def deep_research_evidence(topic: str, country: str = "us") -> dict:
    async with httpx.AsyncClient(base_url="https://api.searchdataapi.com", headers=HEADERS, timeout=15.0) as client:
        # Parallel fan-out: Web SERP + PAA (2 Cr) + Live Google News Wire (2 Cr) = $0.0004 total!
        serp_task = client.get("/v1/search", params={"q": topic, "country": country, "limit": 8})
        news_task = client.get("/v1/news/search", params={"q": topic, "country": country.upper(), "when": "7d", "limit": 6})
        serp_res, news_res = await asyncio.gather(serp_task, news_task)

    serp = serp_res.json()["data"]
    news = news_res.json()["data"]
    return {
        "topic": topic,
        "authoritative_web_sources": [
            {"rank": h["position"], "title": h["title"], "url": h["link"], "snippet": h["snippet"]}
            for h in serp.get("organic_results", [])[:5]
        ],
        "people_also_ask": serp.get("people_also_ask", [])[:4],
        "breaking_news_citations": [
            {"publisher": a.get("source"), "published_utc": a.get("published_at"), "title": a["title"], "url": a["link"]}
            for a in news.get("articles", [])[:5]
        ],
    }

if __name__ == "__main__":
    bundle = asyncio.run(deep_research_evidence("solid state transformer data center"))
    print(f"Grounded {len(bundle['authoritative_web_sources'])} web + {len(bundle['breaking_news_citations'])} news citations")
Workflow 03 · GET /v1/autocomplete + GET /v1/search

Sub-80ms Query Expansion & Intent Discovery Agent

Expand a seed topic via real-time Google Suggest predictions (<80ms, 1 Credit) and fan out parallel SERP + PAA fetches to map complete user search intent before generating content or research reports.

import asyncio
import os
import httpx

API_KEY = os.environ["SEARCHDATAAPI_KEY"]
HEADERS = {"Authorization": f"Bearer {API_KEY}"}

async def expand_and_cluster_intent(seed_prefix: str, country: str = "us") -> dict:
    async with httpx.AsyncClient(base_url="https://api.searchdataapi.com", headers=HEADERS, timeout=15.0) as client:
        # Step 1: <80ms Google Suggest Autocomplete fan-out (1 Credit = $0.0001)
        ac = (await client.get("/v1/autocomplete", params={"q": seed_prefix, "country": country, "limit": 5})).json()["data"]
        expansions = [s["value"] if isinstance(s, dict) else str(s) for s in ac.get("suggestions", [])[:4]]

        # Step 2: Fetch top SERP rankings + PAA questions in parallel for each expanded intent
        serp_responses = await asyncio.gather(
            *[client.get("/v1/search", params={"q": q, "country": country, "limit": 5}) for q in expansions]
        )

    clusters = []
    for query, resp in zip(expansions, serp_responses):
        d = resp.json()["data"]
        clusters.append({
            "intent_query": query,
            "top_urls": [r["link"] for r in d.get("organic_results", [])[:3]],
            "paa_questions": [p["question"] for p in d.get("people_also_ask", [])[:3]],
        })
    return {"seed": seed_prefix, "clusters": clusters}

if __name__ == "__main__":
    print(asyncio.run(expand_and_cluster_intent("best rag framework for")))

Give your AI agents live Google SERP, Trends & News today.

Start free with 3,000 Credits ($0.30) on sign-up — no credit card required — or scale with the $9 never-expiring Indie Starter Pack.