---
title: "The 10 Best Free SERP APIs of 2026"
slug: best-free-serp-apis
date: 2026-10-14T13:13:45+00:00
modified: 2026-10-14T13:13:48+00:00
permalink: https://brightdata.com/blog/web-data/best-free-serp-apis
type: blog
---

[ Blog ](https://brightdata.com/blog "Blog") / [Web Data](https://brightdata.com/blog/web-data)







 [Web Data](https://brightdata.com/blog/web-data)

# The 10 Best Free SERP APIs of 2026

The 10 best free SERP APIs of 2026, ranked by real free searches. Bright Data leads with 5,000 free requests a month, usable from MCP, the CLI or the API.

 21 min read





 [ ](https://brightdata.com/blog/authors/amitesh-anand)

 [Amitesh Anand

Technical Writer

 ](https://brightdata.com/blog/authors/amitesh-anand)





 ![best-free-serp-apis-F-funnel](https://media.brightdata.com/2026/10/best-free-serp-apis-F-funnel.png)





This guide covers the 10 best free SERP APIs of 2026, ranked by how many free searches you actually get. First is the Bright Data [SERP API](/products/serp-api), with 5,000 free requests every month and no credit card. That allowance renews on the first of each month. It works from a hosted MCP server, a command-line tool or a single REST call. So a coding agent can search Google without anyone writing scraping code.

“Free” means very different things across these vendors. Some allowances renew monthly, while others are a one-time sign-up credit. Some APIs charge 5 to 25 credits for one Google search. A big credit number can hide a small search count. This list converts every free plan into real searches and flags credit card rules. It also covers free options that no longer exist, including Google’s own Custom Search JSON API.

In this guide, you will learn:

- How the 10 free SERP APIs compare on free searches, renewal, card rules and price after the free tier.
- Which popular free search options are closed, retired or limited.
- How to use Bright Data’s free SERP API from a coding agent through MCP or the Bright Data CLI.
- What 5,000 free searches a month buy, and what happens when you run out.

## TL;DR: Best Free SERP APIs at a Glance

ToolFree allowanceRenewsCard neededResults fromAfter the free tier**Bright Data**5,000 requests/moMonthlyNoGoogle, Bing + 5 more$1.50/1K PAYG, from $1/1K**Decodo**Up to 5K crawls/moMonthlyNoGoogle, Bing$0.75/1K (/search), plans from $19/mo**Apify**$5 credit/mo (~1,111 pages)MonthlyNoGoogle$2.50/1K pages on Starter ($19/mo)**Oxylabs**$5 credit (~5,000 results)One-timeNoGoogle, Bing$1/1K Google results, plans from $30/mo**SerpApi**250 searches/moMonthlyNoGoogle, Bing + many more$25/mo for 1,000 searches**Serper**2,500 queriesOne-timeNoGoogle only$1/1K, down to $0.30/1K**DataForSEO**$1 credit (~1,666 SERPs)One-timeNoGoogle, Bing + 5 more$0.60/1K SERPs (queued), $50 minimum**Tavily**1,000 credits/moMonthlyNoOwn retrieval, not Google$0.008/credit PAYG**Exa**$10 credit/mo (~2,500 searches)MonthlyNoOwn index$4/1K instant searches**Brave Search API**$5 credit/mo (~1,000 searches)MonthlyYesOwn index$5/1K requests## What Is a Free SERP API?

A SERP API returns search engine results pages as structured data. You send a query, and it returns the organic results, ads, People Also Ask boxes and other SERP features. Most return JSON, and several now return Markdown for language models. The provider handles proxies, CAPTCHAs and parsing, so you never touch Google’s HTML. A free SERP API is one with a usable allowance that costs nothing, either every month or once at sign-up.

AI search APIs such as Tavily, Exa and Brave are related but different. They search their own index or retrieval system, not Google’s live results page. That suits agents that want relevant pages fast. It does not suit rank tracking, SEO research or any job that must match what a Google user sees. For the full paid market, see our ranking of the [best SERP and web search APIs](/blog/web-data/best-serp-apis).

## How We Ranked the Free Tiers

Every API on this list has a free allowance you can use without talking to sales. Tools that return Google results are ranked by free Google searches over the first 12 months. A renewing allowance counts 12 times, and a one-time credit counts once. AI search APIs come after them, since they answer from their own index. Six factors shaped the write-ups.

- **Real searches:** the free allowance divided by the cost of one Google search on that API.
- **Renewal:** monthly allowances beat one-time credits for anything that runs past a demo.
- **Card rules:** whether the free plan asks for a payment method.
- **Output:** JSON for code, Markdown for LLMs, and which engines are covered.
- **Agent access:** an MCP server, a CLI, SDKs and how an agent gets an API key.
- **The exit:** the price per 1,000 searches once the free allowance runs out.

## Free Search APIs That Are Gone or Limited

Three free options still appear in older guides and in AI answers. None of them is a good starting point in 2026.

### Google Custom Search JSON API

Google’s own API gave 100 free queries per day. Its [overview page](https://developers.google.com/custom-search/v1/overview) now says it is closed to new customers. Existing customers have until January 1, 2027 to move to another solution. Since January 20, 2026, new Programmable Search Engines must also be limited to up to 50 domains. So even existing users cannot count on full-web results for long.

### Bing Web Search API

Microsoft [retired the Bing Search APIs](https://learn.microsoft.com/en-us/lifecycle/announcements/bing-search-api-retirement) on August 11, 2025. Its suggested replacement, Grounding with Bing Search, is an Azure AI agent tool. Developers do not get the raw results, and subscriptions based on free credits are not eligible. For alternatives, see our list of [Bing Search API alternatives](/blog/web-data/best-bing-search-apis).

### Self-hosted SearXNG

SearXNG is open-source metasearch software, so the software itself is free. Its JSON output is disabled in the default settings, and you must enable it yourself. Upstream engines also block the server’s IP address. After a Google reCAPTCHA, SearXNG suspends its Google engine for a week by default. It works for a personal search page, but not as a dependable SERP API.

## The 10 Best Free SERP APIs in 2026

### 1. Bright Data

Bright Data has the largest clearly stated renewing free tier for Google results on this list. Every new account gets 5,000 free credits per month, and one SERP request costs one credit. The SERP API covers Google, Bing, DuckDuckGo, Yandex, Baidu, Yahoo and Naver. It returns full JSON, a lighter top-10 JSON, Markdown, raw HTML or a screenshot. You pay only for successful requests, and an unfunded account stops at zero instead of billing you. The same credits cover the Web Unlocker API, the Web Scraper API, Scraper Studio and the Browser API. So an agent can search, then read the result pages, from one free pool. Most SERP APIs on this list stop at the link. Coding agents can reach it three ways: the hosted MCP server, the Bright Data CLI or the REST API. An agent can even create the account itself with the [agent registration protocol](https://brightdata.com/auth.md).

**Key details:**

- **Free tier:** 5,000 credits per month, renewed on the 1st, no credit card. Unused credits do not roll over.
- **Engines:** Google, Bing, DuckDuckGo, Yandex, Baidu, Yahoo and Naver.
- **Output:** full JSON, light JSON with the top 10 organic results, Markdown, HTML and screenshots.
- **Agent access:** hosted MCP server, Bright Data CLI, agent skills, and Python and JavaScript SDKs.
- **Rate limit on the free tier:** 1,000 requests per minute, removed once you add funds.
- **After the free tier:** $1.50 per 1,000 requests pay-as-you-go, down to $1 per 1,000 on the $1,999 plan.

**Pros:**

- Largest clearly stated renewing free allowance for Google results, with no card.
- One free pool covers search and page fetching for agents.
- Seven search engines and Markdown output for LLMs.
- Hard stop at zero on unfunded accounts, so no surprise bills.
- An agent can sign up, get an API key and a ready SERP zone without a web form.

**Cons:**

- Google requests sometimes fail on a CAPTCHA or a render timeout. They are not billed, but your code should retry.
- Proxy products are not part of the monthly free credits.
- Unused free credits do not carry over to the next month.

⭐ **Review score:** G2 4.7/5 (346 reviews)

#### Get 5,000 free SERP API requests every month

Google, Bing and five more engines as JSON or Markdown. No credit card. Use it from the MCP server, the CLI or one REST call.

 [
 Start free
 ](https://brightdata.com/cp/start)
 [
 See SERP API pricing
 ](https://brightdata.com/pricing/serp)

### 2. Decodo

Decodo, formerly Smartproxy, now has a $0 monthly plan for its Web Scraping API. The pricing page lists it as up to 5,000 crawls or 10,000 pages scraped per month, with no card. Its google\_search template returns parsed Google results and costs $1 per 1,000 on the default premium proxy pool. The newer Fast Search API costs $0.75 per 1,000 and promises sub-second organic results. The free plan does not state its budget in searches, and the Fast Search docs cite 10,000 requests instead.

**Key details:**

- **Free tier:** $0 per month, up to 5,000 crawls or 10,000 pages, no credit card.
- **Engines:** Google (Search, AI Mode, Hotels, Autocomplete) and Bing.
- **Output:** JSON, HTML, Markdown, CSV and PNG, depending on the template.
- **Agent access:** hosted MCP server, CLI, agent skills, and Python, TypeScript and Go SDKs.
- **After the free tier:** plans from $19 per month.

**Pros:**

- Renewing free plan in the same range as Bright Data’s.
- Search and general scraping share one plan.
- Large G2 review base across Decodo’s products.

**Cons:**

- The free allowance is stated in crawls and pages, not searches, and two pages give different numbers.
- JavaScript rendering raises the Google price to $1.50 per 1,000.

⭐ **Review score:** G2 4.6/5 (751 reviews)

### 3. Apify

Apify is a marketplace of scrapers called Actors, and its free plan includes $5 of usage every month. The official Google Search Results Scraper charges $4.50 per 1,000 result pages on the free plan. That works out to about 1,111 Google result pages a month, each holding roughly 10 results. Runs can be scheduled, and results export as JSON, CSV or Excel. Add-ons such as AI Overviews and AI Mode cost extra and drain the $5 quickly.

**Key details:**

- **Free tier:** $5 of platform usage per month, no credit card. Unused credit expires at the end of the cycle.
- **Engine:** Google, with optional AI Mode, ChatGPT, Perplexity, Copilot and Gemini add-ons.
- **Output:** Apify dataset exported as JSON, CSV, Excel and more.
- **Agent access:** Apify MCP server, API, CLI and SDKs.
- **After the free tier:** Starter at $19 per month, which drops the Actor price to $2.50 per 1,000 pages.

**Pros:**

- Renewing monthly credit with no card.
- No-code scheduling and exports for non-developers.
- Large review base for the Apify platform.

**Cons:**

- Priced per result page, so deep searches use the credit fast.
- AI Mode costs $200 per 1,000 on the free plan.

⭐ **Review score:** G2 4.7/5 (653 reviews)

### 4. Oxylabs

Oxylabs gives every new account a one-time $5 welcome credit with no payment details. The credit stays valid indefinitely as long as you remain on the free plan. Google results without JavaScript rendering cost $1 per 1,000, so the credit covers about 5,000 results. With rendering, the price rises to $1.35 per 1,000, or about 3,700 results. The credit is shared at the team level and works across all Oxylabs products.

**Key details:**

- **Free tier:** one-time $5 credit, no card, no expiry while on the free plan.
- **Engines:** Google and Bing.
- **Output:** parsed JSON, HTML and Markdown.
- **Agent access:** official MCP server and agent skills.
- **After the free tier:** plans from $30 per month.

**Pros:**

- The welcome credit does not expire while you stay on the free plan.
- About 5,000 Google results is a generous one-time test.
- Strong G2 standing across 422 reviews.

**Cons:**

- The credit never renews.
- One $5 credit per team, not per user.

⭐ **Review score:** G2 4.5/5 (422 reviews)

### 5. SerpApi

SerpApi has the widest SERP feature coverage on this list, with dozens of Google result types and many other engines. Its free plan includes 250 searches per month, with throughput of 50 searches per hour. One search costs one credit, whatever the number of results. Repeating the same search within an hour is served from cache and costs nothing. Output comes as JSON, raw HTML or Markdown for LLMs.

**Key details:**

- **Free tier:** 250 searches per month. Signing up asks for no payment details.
- **Engines:** Google, Bing, Baidu, Yahoo, Yandex, DuckDuckGo, Naver and more.
- **Output:** JSON, HTML and Markdown.
- **Agent access:** hosted MCP server and official libraries in 10 languages.
- **After the free tier:** $25 per month for 1,000 searches, or $75 for 5,000. No pay-as-you-go.

**Pros:**

- Broadest coverage of SERP features and engines.
- Free cached repeats within an hour.
- High G2 score, 4.8 across 29 reviews.

**Cons:**

- 250 searches is a small monthly allowance.
- Paid plans cost $15 to $25 per 1,000 searches at the entry level.
- The U.S. Legal Shield starts on the $150 Production plan.

⭐ **Review score:** G2 4.8/5 (29 reviews)

### 6. Serper

Serper is a Google-only search API built around low prices. New accounts get 2,500 free queries, with no credit card. The vendor does not say the free queries renew, so treat them as a one-time credit. After that, credits are prepaid, from $50 for 50,000 down to $0.30 per 1,000 at the largest tier. Paid credits expire after six months, and only successful responses use credits.

**Key details:**

- **Free tier:** 2,500 queries once, no credit card.
- **Engine:** Google only: Search, Images, News, Maps, Places, Videos, Shopping, Scholar and Patents.
- **Output:** JSON.
- **Speed claim:** the vendor quotes 1 to 2 seconds per query.
- **After the free tier:** $1 per 1,000 at the entry top-up, falling to $0.30 per 1,000.

**Pros:**

- Large one-time free credit.
- Very low prices at high volume.
- Pay-as-you-go top-ups, no subscription.

**Cons:**

- Google only.
- No renewing free allowance.
- Paid credits expire after six months.

### 7. DataForSEO

DataForSEO is a pay-as-you-go SEO data API with a SERP API for Google, Bing, Yahoo, Baidu, Naver, Seznam and YouTube. New accounts get a $1 trial credit and a free sandbox, with no card. The balance has no expiry date. The standard queue costs $0.0006 per SERP of 10 results and returns in about five minutes. At that rate, $1 covers about 1,666 SERPs, or 500 in live mode at $0.002 each. It suits batch keyword jobs better than real-time agent calls.

**Key details:**

- **Free tier:** one-time $1 credit plus a sandbox, no card, no expiry.
- **Engines:** Google, Bing, Yahoo, Baidu, Naver, Seznam and YouTube.
- **Output:** parsed JSON or raw HTML.
- **Agent access:** official remote MCP server.
- **After the free tier:** $0.60 per 1,000 SERPs in the standard queue, with a $50 minimum payment.

**Pros:**

- Very low per-SERP price in the queued mode.
- Balance never expires.
- Broad SEO data beyond SERPs.

**Cons:**

- The cheap rate needs a five-minute queue.
- Deeper result pages and advanced operators multiply the cost.

⭐ **Review score:** G2 4.4/5 (17 reviews)

### 8. Tavily

Tavily is a search API built for AI agents, now part of Nebius. Its free Researcher plan includes 1,000 API credits per month, with no credit card. A basic search costs one credit, and an advanced search costs two. Results come back ranked and scored, with extracted page content and an optional generated answer. Tavily does not return Google’s results page, so it is not a fit for rank tracking or SEO work.

**Key details:**

- **Free tier:** 1,000 credits per month, reset on the 1st, no credit card.
- **Results:** Tavily’s own web retrieval, with extracted content.
- **Output:** JSON.
- **Agent access:** hosted MCP server, including a rate-limited keyless mode, plus Python and JavaScript SDKs.
- **After the free tier:** $0.008 per credit pay-as-you-go, or $30 per month for 4,000 credits.

**Pros:**

- Answer-ready results with page content included.
- Keyless MCP access for quick experiments.
- Renewing free allowance.

**Cons:**

- No Google SERP data.
- At $8 per 1,000 basic searches, paid usage costs several times more than SERP APIs.

⭐ **Review score:** G2 4.5/5 (3 reviews)

### 9. Exa

Exa runs its own web index and search models, built for semantic, meaning-based queries. The free tier gives $10 of credit every month, plus a one-time $10 bonus, with no payment method. Instant searches cost $4 per 1,000, so the monthly credit covers about 2,500 of them. The default auto mode costs $7 per 1,000, which cuts that to about 1,400. Like Tavily, it does not return Google’s results page.

**Key details:**

- **Free tier:** $10 credit per month plus a $10 sign-up bonus, no payment method.
- **Results:** Exa’s own index.
- **Output:** JSON with optional page text, highlights and summaries.
- **Agent access:** hosted MCP server, Python and JavaScript SDKs, and agent skills.
- **After the free tier:** $4 per 1,000 instant searches, $7 per 1,000 in auto mode.

**Pros:**

- Strong at semantic and similarity queries.
- Renewing $10 credit with no card.
- Optional page contents in the same call.

**Cons:**

- No Google SERP data.
- The free search count depends on the search mode.

⭐ **Review score:** G2 4.5/5 (1 review)

### 10. Brave Search API

The Brave Search API queries Brave’s independent web index, the one behind Brave Search. The Search plan includes $5 of free credit each month, which covers about 1,000 searches at $5 per 1,000. Brave requires a credit card for the free credit as an anti-fraud check, and says it will not charge it. Its terms forbid storing search results beyond transient use. They also forbid using results to train, fine-tune or benchmark AI models.

**Key details:**

- **Free tier:** $5 credit per month, credit card required.
- **Results:** Brave’s own index.
- **Output:** JSON for web, news, images and video, plus an LLM context endpoint.
- **Agent access:** official MCP server.
- **After the free tier:** $5 per 1,000 requests.

**Pros:**

- Independent index, not a Google or Bing reseller.
- LLM context endpoint returns pre-extracted content.
- See our [Bright Data vs. Brave Search API](/blog/ai/bright-data-vs-brave-search-api) comparison for a deeper look.

**Cons:**

- Needs a credit card even for the free credit.
- Terms restrict storing results and using them for AI training.
- No Google SERP data.

## Free Tiers That Did Not Make the Cut

Several other SERP APIs advertise a free plan. Once you convert credits into Google searches, the allowances are small or unclear.

ToolFree allowance as statedReal Google searchesWhy it is not on the listHasData1,000 credits per month100 full SERPs per month10 credits per full SERP requestSearchApi100 free requests100Renewal is not statedSerpstack100 searches per month100 per monthFree plan is for non-commercial use onlyZenserp50 searches per month50 per monthGoogle web search only on the free planScraperAPI7-day trial with 5,000 credits, then 1,000 credits per month200 in the trial, then 40 per monthEach Google request costs 25 creditsScrapingdog100 credits20Each Google search costs 5 creditsSerpHouse4,000 creditsAbout 400Pricing page and terms disagree on monthly renewal versus a 7-day trialLinkup$20 credit per monthAbout 4,000 searchesOwn index, and only for eligible professional email addresses## How to Use Bright Data’s Free SERP API From a Coding Agent

Coding agents such as Claude Code, Cursor and Codex can use the free tier in three ways. The MCP server gives the agent search as a native tool. The CLI gives it a shell command it can run and parse. The REST API is for code you ship in your own app. All three draw from the same 5,000 monthly credits.

### Step 1: Create a free account

A person can sign up at [brightdata.com](https://brightdata.com/cp/start) with no credit card and copy an API key from account settings. An agent can also do it without a web form, using the [auth.md registration protocol](https://brightdata.com/auth.md). It posts the user’s email address, asks the user for the 6-character code they receive, then claims the account. The claim returns an API key plus three ready zones: `agent_serp`, `agent_unlocker` and `agent_browser_api`. Our guide to [agentic registration](/blog/ai/agentic-registration) walks through the full flow.

Store the key in an environment variable, never in a prompt or a committed file:

  ```bash
export BRIGHTDATA_API_KEY="YOUR_API_KEY"
```

### Option 1: Connect the Bright Data MCP server

The [Bright Data MCP server](/ai/mcp-server) is the most direct route for an agent. The hosted server needs no install, and you choose which tools it exposes. Listing only `search_engine` and `scrape_as_markdown` keeps tool definitions to about 360 tokens. The default set of 11 tools takes about 2,000. In Claude Code, one command adds the server and sends the key as a header instead of in the URL:

  ```bash
claude mcp add --transport http brightdata \
  "https://mcp.brightdata.com/mcp?tools=search_engine,scrape_as_markdown" \
  --header "Authorization: Bearer $BRIGHTDATA_API_KEY"
```

Now ask the agent a question that needs fresh results, such as “search Google for FastAPI background task patterns and summarize the top three pages.” It calls `search_engine` for the SERP, then `scrape_as_markdown` to read the pages. Each call uses one credit from the free pool. The hosted server also wraps fetched content in markers that label it as untrusted web data. That helps the agent treat instructions hidden in a web page as data, not commands.

Using Cursor or Codex, or more than one agent? The Bright Data CLI writes the MCP configuration for you. This adds a local server to each agent you name, with your API key in its environment:

  ```bash
bdata add mcp --agent claude-code,cursor,codex --global
```

MCP clients that support OAuth can also sign in through the browser instead of using a key. See the [MCP OAuth guide](https://docs.brightdata.com/products/mcp-server/remote/oauth) for details. For a deeper look at token budgets, read [lightweight SERP grounding with Bright Data MCP](/blog/ai/lightweight-serp-grounding-with-bright-data-mcp). Our ranking of the [best MCP servers for Claude Code](/blog/ai/best-mcp-servers-for-claude-code) covers what to pair it with.

### Option 2: Search from the terminal with the Bright Data CLI

The Bright Data CLI suits agents that work through a shell, and it is handy for quick checks yourself. Install it from npm, then sign in once. `bdata login` opens a browser, and `bdata login --device` works over SSH or in a headless sandbox. Login also creates the zones the CLI needs, so search works right away.

  ```bash
npm install -g @brightdata/cli
bdata login
bdata search "fastapi background tasks"
```

By default, `bdata search` prints Google results as a table. Add `--json` for parsed JSON an agent can read, or `-o results.json` to write a file. Switch engines with `--engine bing` or `--engine yandex`, and localize with `--country` and `--language`. Use `--type news`, `images` or `shopping` for other result types, and `--page` to paginate. Check your account balance with `bdata budget balance`.

To make an agent reach for the CLI on its own, install the Bright Data agent skills. They teach the agent when to search, when to scrape and which command to run:

  ```bash
npx skills add brightdata/skills
```

The same skills work in Claude Code, Cursor, Codex and other agents. Our list of the [best CLI tools for Codex](/blog/ai/best-cli-tools-for-codex) shows the CLI next to other agent-friendly tools.

### Option 3: Call the REST API from your own code

For an app or a pipeline, call the SERP API directly. Send a POST to `https://api.brightdata.com/request` with your zone and a normal Google search URL. Add `data_format: parsed_light` to get the top 10 organic results as JSON. Use `brd_json=1` in the URL for the full parsed page, or `data_format: markdown` for LLM-ready text. This script uses the `agent_serp` zone from agent registration; set `BRIGHTDATA_SERP_ZONE` to use your own.

  ```python
import os
import sys
import time
from urllib.parse import quote_plus

import requests

API_KEY = os.environ["BRIGHTDATA_API_KEY"]
ZONE = os.environ.get("BRIGHTDATA_SERP_ZONE", "agent_serp")
RETRYABLE = {"expect_element", "expect_body", "captcha", "timeout"}

def google(query, country="us", attempts=4):
    """Return Google's top organic results as a list of dicts."""
    url = f"https://www.google.com/search?q={quote_plus(query)}&hl=en&gl={country}"
    for attempt in range(1, attempts + 1):
        r = requests.post(
            "https://api.brightdata.com/request",
            headers={"Authorization": f"Bearer {API_KEY}"},
            json={"zone": ZONE, "url": url, "format": "raw", "data_format": "parsed_light"},
            timeout=90,
        )
        error = r.headers.get("x-brd-error-code")
        if r.ok and r.text and not error:
            return r.json()["organic"], attempt
        if error not in RETRYABLE and r.status_code not in (429, 502, 503):
            raise RuntimeError(f"HTTP {r.status_code} {error}: {r.headers.get('x-brd-error')}")
        time.sleep(2 ** attempt)
    raise RuntimeError(f"No results after {attempts} attempts ({error})")

if __name__ == "__main__":
    results, attempts = google(" ".join(sys.argv[1:]))
    print(f"{len(results)} results ({attempts} attempt(s))")
    for item in results[:5]:
        print(f"{item['global_rank']:>2}. {item['title'][:60]}")
        print(f"    {item['link']}")
```

One detail matters for reliable code. Sometimes Google answers with a CAPTCHA or a page that never finishes loading. The API can then return HTTP 200 with an empty body. The reason is in the `x-brd-error-code` header, with values such as `captcha` or `expect_element`. Those failed requests are not billed, so the script retries them with a growing delay. Also skip the `num` parameter: the API removes it and warns you in an `x-brd-serp-warn` header.

## What 5,000 Free Searches a Month Buy

Five thousand requests go further than they sound, as long as you match the plan to the job. Each search or page fetch costs one credit.

WorkloadCredits per runFree runs per monthCoding agent research task: 10 searches and 3 page reads13About 384 tasksDaily rank check for 160 keywords160 per dayEvery day of a 31-day monthRAG grounding: 1 search per user question15,000 questionsWeekly SERP snapshot of 1,000 keywords1,000 per week4 weekly snapshots, with room to spareAn unfunded account stops at zero, so a runaway agent loop cannot run up a bill. Free accounts are also limited to 1,000 requests per minute, which is plenty for agents and small pipelines. The control panel shows your remaining free credits and the next renewal date. Credits reset on the first of each month and do not roll over.

## What Happens After the Free Tier

When the free credits run out, you can wait for the next month or add funds. Pay-as-you-go costs $1.50 per 1,000 requests, with no commitment and no minimum. Monthly [SERP API plans](/pricing/serp) start at $499 for 380,000 requests and fall to $1 per 1,000 on the $1,999 plan. At the time of writing, Bright Data also matches a first deposit dollar for dollar, up to $500. Turn on auto-recharge if an agent or pipeline must never stop.

Note that the monthly free credits stop if you move to a committed subscription plan. Credits left at the time of the upgrade stay usable. For context, SerpApi’s entry plans work out to $15 to $25 per 1,000 searches. Serper, Oxylabs and DataForSEO cost less per search, but none of them renews a free allowance.

## Conclusion

The best free SERP API depends on how long you need it and what results you need. For ongoing Google results, Bright Data has the largest clearly stated free allowance: 5,000 requests a month, no card. Its MCP server and CLI also make it the quickest to wire into a coding agent. Decodo is the closest free alternative, and Oxylabs and Serper are good one-time tests. If you want AI-ranked pages rather than Google’s results page, Tavily and Exa have useful free tiers.

Start with the free allowance, measure your real query volume for a month, then decide. Going further? Read our guides to the [best MCP servers for Codex](/blog/ai/best-mcp-servers-for-codex), [Google search URL parameters](/blog/web-data/google-search-url-parameters) and the [best free web scrapers](/blog/web-data/best-free-web-scrapers).

## Frequently Asked Questions

### Is there a free SERP API?

Yes. Bright Data gives 5,000 free SERP API requests every month with no credit card. Decodo, Apify and SerpApi also have renewing free plans, and Oxylabs, Serper and DataForSEO give one-time credits.

### Is the Google Search API free?

Not for new users. Google’s Custom Search JSON API offered 100 free queries per day, but it is closed to new customers. Existing customers must move to another solution by January 1, 2027. Third-party SERP APIs are the practical route to Google results.

### Which free SERP API gives the most searches?

For Google results that renew every month, Bright Data’s 5,000 requests is the largest allowance stated in searches. Decodo lists up to 5,000 crawls or 10,000 pages a month. For a one-time credit, Oxylabs’ $5 covers about 5,000 Google results.

### Do free SERP APIs require a credit card?

Most do not. Bright Data, Decodo, Apify, Oxylabs, SerpApi, Serper, DataForSEO, Tavily and Exa all start without a card. Brave Search API requires a card for its free monthly credit.

### What is the best free search API for AI coding agents?

Bright Data is a strong fit. It has a hosted MCP server, a CLI with agent skills, and a sign-up protocol an agent can complete itself. Its 5,000 monthly credits cover both searches and page fetches.

### What happens when the Bright Data free tier runs out?

Without funds, requests stop until the credits renew on the first of the next month. There is no surprise bill. With funds added, usage continues at $1.50 per 1,000 requests pay-as-you-go.

### Is it legal to scrape Google search results?

Collecting publicly available search results is generally legal in the US and EU. Avoid personal data you have no basis to process, and follow privacy laws such as GDPR. Check with legal counsel for your specific use case.



Contact usStart for free

No credit card required











 [ ](https://www.linkedin.com/in/amitesh1208)

Amitesh Anand

 Technical Writer





Amitesh Anand is a developer advocate and technical writer sharing content on AI, software, and devtools, with 10k followers and 400k+ views.



Expertise

  AI Agents   Python   Devtools



 [ View all articles ](https://brightdata.com/blog/authors/amitesh-anand)











 Table of Contents







Dedicated Scraper APIs &amp; No-Code Scrapers

Over 1000 scrapers for all popular domains. Simplify your web scraping.

[See pricing](/pricing/web-scraper "See pricing")

Just want data? Skip scraping.

Hundreds of ready-to-use datasets from all popular domains.

[See pricing](/pricing/datasets "See pricing")







 [ ](https://news.ycombinator.com/submitlink?t=The+10+Best+Free+SERP+APIs+of+2026&u=https://brightdata.com/blog/web-data/best-free-serp-apis) [ ](https://www.linkedin.com/shareArticle?mini=true&title=The+10+Best+Free+SERP+APIs+of+2026&url=https://brightdata.com/blog/web-data/best-free-serp-apis) [ ](http://www.reddit.com/submit?title=The+10+Best+Free+SERP+APIs+of+2026&url=https://brightdata.com/blog/web-data/best-free-serp-apis)







##  You might also be interested in

 [ ](https://brightdata.com/blog/products-updates/digitalocean-action-gateway-bright-data "The Shortest Path From Web Data to a Running Agent")

 [Products updates





Bright Data





### The Shortest Path From Web Data to a Running Agent

Bright Data and DigitalOcean give agents live web search, page access, and structured data through DigitalOcean Action Gateway.



 13-Oct-2026

 3 min read

 ](https://brightdata.com/blog/products-updates/digitalocean-action-gateway-bright-data)

 [ ](https://brightdata.com/blog/web-data/best-booking-scrapers "The 10 Best Booking.com Scrapers of 2026")

 [Web Data





Dvir Sharon

Growth Marketing Manager





### The 10 Best Booking.com Scrapers of 2026

Compare the 10 best Booking.com scrapers of 2026 on verified pricing, data depth, and unblocking, from scraper APIs to no-code tools, led by Bright Data.



 08-Oct-2026

 18 min read

 ](https://brightdata.com/blog/web-data/best-booking-scrapers)

 [ ](https://brightdata.com/blog/ai/lightweight-serp-grounding-with-bright-data-mcp "The Lean RAG Engine: How to Use Bright Data MCP for Lightweight SERP Grounding")

 [AI





Daniel Shashko

Web Data &amp; AI Expert





### The Lean RAG Engine: How to Use Bright Data MCP for Lightweight SERP Grounding

How to set up Bright Data MCP for lightweight SERP grounding: one search tool, a search-first prompt and measured token costs per tier.



 07-Oct-2026

 14 min read

 ](https://brightdata.com/blog/ai/lightweight-serp-grounding-with-bright-data-mcp)
