Get Google results as JSON or Markdown

The Search API runs a Google search and returns the results as JSON or Markdown. One endpoint covers web results, news, shopping, places, autocomplete, Google Scholar, and AI Mode.

The same API key works for the Fetch API, which reads a web page as Markdown, and for the MCP server, which gives AI agents both tools.

Before you begin

You need:

• a PacketStream account with funds in its balance. Searches draw from the same prepaid balance as proxy traffic;
• an API key. Create one on the Search API page of the dashboard. The key is shown once, so copy it when it appears.

Load the key into your shell without writing it into the command history:

read -r -s -p 'PacketStream API key: ' PACKETSTREAM_API_KEY
printf '\n'
export PACKETSTREAM_API_KEY

Make your first request

curl -s https://api.packetstream.io/v1/search \
  -H "Authorization: Bearer $PACKETSTREAM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query": "coffee grinders", "country": "us"}'

The response lists up to 10 web results, each with position, title, url, and snippet. It also includes any rich blocks Google showed, such as an AI overview. To search news, shopping, or another Google surface, set type.

Python

import os
import requests

response = requests.post(
    "https://api.packetstream.io/v1/search",
    headers={"Authorization": f"Bearer {os.environ['PACKETSTREAM_API_KEY']}"},
    json={"query": "coffee grinders", "type": "news"},
    timeout=35,
)
response.raise_for_status()
for result in response.json()["results"]:
    print(result["position"], result["title"], result["url"])

JavaScript

Run this on a server, never in a browser, so the key stays private.

const response = await fetch("https://api.packetstream.io/v1/search", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.PACKETSTREAM_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({ query: "coffee grinders", type: "news" }),
  signal: AbortSignal.timeout(35_000),
});
const payload = await response.json();
if (!response.ok) throw new Error(`${payload.error.code}: ${payload.error.message}`);
console.log(payload.result_count, payload.results[0]);

The server finishes or gives up within 25 seconds, so use a client timeout of at least 35 seconds.

Pricing

• Each successful search costs $0.0025, which is $2.50 per 1,000, for every request type and option.
• Failed requests are free: validation errors, rate limits, balance errors, and upstream failures. One rare exception is described under billing.
• A successful search that finds no results is still billed, because the search ran.
• A search needs at least $0.0025 in your balance before it starts.

The dashboard shows daily usage and spend. GET /v1/account returns your balance and usage counts for free.

Authentication

Send the key in an Authorization: Bearer header on every request. Keys start with ps_, followed by 64 hexadecimal characters.

• An account has one active key. It works for the Search API, the Fetch API, and the MCP server.
• Rotating the key on the dashboard revokes the old key immediately. Revoking it stops all API access until you create a new one.
• A missing, malformed, revoked, or unknown key returns 401 unauthorized.
• Keep the key out of URLs, source control, and code that runs in a browser. The API sends no CORS headers, so web pages cannot call it directly.

Base URL and endpoints

Send every request over HTTPS to https://api.packetstream.io.

GET /v1/search takes the same fields as query parameters. Prefer POST for sensitive queries: a GET puts the query in the URL, where proxies and logs along the way can record it.

Privacy

For billing and support, the API records each request’s ID, type, outcome, latency, and charge. It does not store your query text or the results.