Frequently Asked Questions

Pay-as-you-go by product category. Your pricing tier is locked in by your largest single deposit and never changes:

Tier	Google SERP	News	Images	Unlocks at
Default	$0.60 / 1K pages	$0.20 / 1K	$0.35 / 1K	$10 minimum deposit
Growth	$0.06 / 1K pages	$0.02 / 1K	$0.32 / 1K	One-time $100 deposit
Scale	$0.03 / 1K pages	$0.01 / 1K	$0.25 / 1K	One-time $500 deposit

Web Search bills in units of the published rate: Quick Search (/api/search/quick) is one unit per call whatever the page count, and Deep Search (/api/search) is one unit per page requested, up to 10 pages. So at Scale that's $0.03 per 1,000 SERP pages — the cheapest published per-page rate of any SERP API.

Google Maps uses separate endpoint prices and smaller discounts: Quick $15/$13.50/$10.50 per 1K calls, Place $1.50/$1.35/$1.05, and Reviews $3/$2.70/$2.10 across Default/Growth/Scale. Deep has two flat price bands, chosen by the detail parameter. Every Deep call returns up to 100 ranked places whatever you ask for; detail only decides how many of them are opened and fully enriched, and the rest arrive as core-only records that still carry name, address, coordinates, rating, review count, phone and website. A detail of 0 to 20 is the standard band and the default: $0.0150 per call, which is $15/$13.50/$10.50 per 1K calls, $0.15 per 1,000 places returned, or $0.75 per 1,000 fully-enriched places. A detail of 21 to 100 is the premium band, flat: $0.0750 per call, which is $75/$67.50/$52.50 per 1K calls, $0.75 per 1,000 places returned, and at detail=100 also $0.75 per 1,000 fully-enriched places. The step between the bands is a cliff on purpose — detail=20 bills $0.0150 and detail=21 bills $0.0750, five times as much for one extra enriched place — so pick your detail before you build. Synchronous Deep requests are capped at a detail of 30, which is a performance cap rather than a pricing one: 21 to 30 works synchronously and simply pays the premium rate. Deeper pulls, up to 100, are submitted as async jobs at GET /api/maps/search/async and billed at the same flat $0.0750 per call.

Yes. Every new account gets Free to start, shared across every endpoint — Web Search, News, Images, Videos, AI Rank, LinkedIn, Maps, YouTube, and Instagram — no credit card required.

The SERP allowance covers /api/search and /api/search/quick. The Maps allowance covers all four Maps endpoints — Quick, Deep, Place and Reviews.

Free accounts run at the free rate-limit bracket: 1 concurrent request, 10/min, 100/hour, 500/day. Limits increase as soon as you add credits — see What are the rate limits? below.

It depends on the endpoint. /api/search/quick is one published web-search unit per call — any page count, any tier, any balance. /api/search is one published web-search unit per page requested, up to 10 pages per call, at every tier and every balance: a 1-page call is 1 unit and a 10-page call is 10.

/api/search/quick always uses one published Web category unit per request, whatever your tier or balance. /api/search uses one published Web category unit per page requested, up to 10 pages per call, at every tier and every balance — a 1-page call is 1 unit and a 10-page call is 10.

Scale tier Web pricing is $0.03 / 1K pages.

No. Pure pay-as-you-go. No monthly minimum, no recurring charge, no per-seat fee. Top up credits when you want; they don't expire.

Unlike SerpApi, Serper.dev, Zenserp, and most SERP APIs that gate features behind tiered subscriptions, every endpoint and feature is available from day one.

Purchased credits never expire. Promotional bonus credits expire 30 days after they are granted; only the bonus portion expires.

Your pricing tier is locked in by your largest single deposit. A one-time $100 deposit locks Growth pricing (10× off core categories and 10% off Maps) forever. A one-time $500 deposit locks Scale pricing (20× off core categories and 30% off Maps) forever. It never downgrades and never expires, even if your balance spends down.

Note: the lock is set by a single deposit, not cumulative top-ups. A user who deposits $50 + $50 stays on Default; a single $100 deposit locks Growth forever.

We accept all major credit and debit cards (Visa, Mastercard, American Express) through our secure payment partner Dodo Payments. All transactions are SSL encrypted and processed securely.

Because credits are delivered instantly, purchases are generally non-refundable. We make exceptions for technical failures, duplicate charges, unauthorized transactions, or service discontinuation. Submit refund requests within 7 days of purchase via info@apiserpent.com. See the full Refund Policy.

Five engines, one consistent JSON shape:

• Google — engine=google
• Bing — engine=bing
• Yahoo — engine=yahoo
• DuckDuckGo — engine=ddg
• Brave — engine=brave

Plus dedicated AI Ranking endpoints (Claude, ChatGPT, Gemini, Perplexity) and Social Media APIs (YouTube, Instagram).

Yes, on eligible Deep Search calls. Add include_aio=true to /api/search with engine=google, engine=yahoo or engine=brave to request AI Overview fields when available. It is off by default so organic SERP calls stay fast and predictable.

If the AI Overview is unavailable, the response still returns the rest of the SERP data with aiOverview set to null. There is no separate "AI Overview API" endpoint.

Pixel position is the y-coordinate (in pixels, from the top of the rendered SERP) of any element on the search results page. Add pixel_position=true to /api/search (Deep Search) and every item — organic results, AI Overview, People Also Ask, featured snippet, knowledge panel, ads, shopping, videos, and local pack — comes back with two sibling fields: an integer pixel_position (the y-coordinate) and a pixel_box object containing the full bounding rectangle { x, y, w, h }.

Currently in BETA, supported on all four engines (Google, Yahoo, Bing, DuckDuckGo) on desktop, US locale. Quick Search does not support pixel positions on any engine. /api/search/quick still accepts pixel_position=true and returns no error; the response carries metadata.pixelPositionUnavailable: 'not_supported_on_quick' and a metadata.pixelPositionHint pointing you to /api/search. See the full spec →

No. Pixel position uses the same per-call price as the underlying /api/search Deep Search call across all four engines. A request with pixel_position=true costs the same as one without — and you get the full pixel_box geometry (x, y, w, h) in the same call. Available on all paid tiers; free-tier responses include metadata.pixelPositionUnavailable: 'free_tier' instead of pixel data.

Rank tells you a result is #1 organic. Pixel position tells you the user has to scroll past 1,800 px of AI Overview, ads, and PAA boxes to actually see it. Two queries with the same rank can have wildly different real visibility and CTR. Pixel position is the metric that catches every SERP layout change Google rolls out.

Read the case study: Pixel Position vs Rank: Why Modern SEO Tools Need Both.

On Google Deep Search (/api/search) results, Serpent extracts:

• Organic results (~10/page, up to 10 pages)
• AI Overview (AIO) — optional on Deep Search
• Featured Snippet (position zero)
• People Also Ask (PAA)
• Related searches
• Inline videos
• Sponsored / ads

Quick Search (/api/search/quick) is organic-only on engine=google; on Yahoo, Bing, DuckDuckGo and Brave it still returns that engine’s supported feature blocks. Use Deep Search for richer Google SERP features, and add include_aio=true when you need AI Overview fields. Video Search also has a dedicated endpoint.

Yes. Pass country=us, country=uk, etc. and language=en for language. We support 112 country codes — see GET /api/countries for the full list. City/locality workflows should use Local or Maps endpoints rather than organic Web Search.

Up to 10 pages in a single /api/search call (~100 results), all paginated for you and merged into one response.

Quick Search (/api/search/quick) accepts num= up to 100: one call, one charge, up to 100 organic results, paginated for you, speed-optimized and best-effort at the high end (num=50 and below reliably returns the full set). One call, one charge is always true on Quick Search. Deep Search is billed per page requested — one unit per page, up to 10 pages — so a 10-page call is 10 units. See "Do you charge per call or per page?" above.

Yes. If an integration repeatedly sends requests that fail because the service is temporarily unavailable, the API may return a temporary 429 cooldown with a Retry-After header. Validation errors, auth errors, billing errors, and normal rate-limit responses are handled separately.

site: queries can return organic SERP results, but Serpent is primarily built for retrieving SERP rankings and result pages. Treat site: output as search-result data, not as a guaranteed indexing audit or proof that a URL is indexed.

In September 2025 Google removed the num=100 query parameter, capping a single SERP page at 10 results. DIY scrapers now have to paginate — start=0, start=10, start=20… — and stitch the results.

Serpent's /api/search endpoint paginates and merges for you. One call, up to ~100 results, all in a single JSON response. No client-side pagination logic, no extra requests. Deep Search bills one unit per page requested, so a full 10-page call is 10 units of the published Web rate.

Quick Search (/api/search/quick) qualifies too, for organic results: one call, one flat charge, num= up to 100 organic results, paginated for you and best-effort at the high end — at the same published Web rate, one unit per Quick call.

Real-time. Every call hits Google fresh — no cached responses, no queue delay. This is unlike DataForSEO Standard Queue (~5-minute delay) or Bright Data's "discovered" tier. If you're rank-tracking on a schedule or building real-time agents, you get current data on every call. Competitor queue behaviour checked on 10 September 2026.

Desktop SERPs only at this time. Mobile SERP rendering and mobile pixel positions are on the roadmap. Reach out at info@apiserpent.com if mobile is a hard requirement for your use case — it helps us prioritize.

AI Ranking measures how visible your brand or URL is when users ask Claude, ChatGPT, Gemini, or Perplexity a question. Instead of reading a search results page, we ask the AI assistant a query and analyze the answer for citations, mentions, and ranking position.

Use it to track AI search visibility ("AEO" — Answer Engine Optimization) the same way you'd track Google rankings. Endpoints: /api/ai/rank/claude, /api/ai/rank/chatgpt, /api/ai/rank/gemini, /api/ai/rank/perplexity.

SerpApi's entry plan is $25/month for 1,000 searches — that's $25 per 1,000 calls. The rate improves with volume: $9.17 per 1,000 on the $275/month Big Data plan, down to about $1.96 per 1,000 on their largest published Cloud plan ($106,050/month for 54M searches), which is the cheapest rate they publish. Read from their own pricing page on 10 September 2026.

Serpent at the Default tier is $0.60 / 1K calls, pay-as-you-go, no subscription — about 40× cheaper than SerpApi's entry plan. At Scale tier (with a one-time $500 deposit) it's $0.03 / 1K calls: over 800× cheaper than their entry plan, and still about 65× cheaper than the $1.96 rate that needs a six-figure monthly spend.

Feature parity: both extract Google AIO, PAA, Featured Snippets, ads, organic, and shopping blocks from the web SERP. SerpApi has a wider engine catalog (80+ engines including niche verticals); Serpent focuses on the five engines that matter for ranking and bundles AI Ranking endpoints (Claude / ChatGPT / Gemini / Perplexity) that SerpApi does not offer.

DataForSEO's Google SERP API is $0.60 per 1K calls in the Standard Queue (~5 minute delay) and $2 per 1K calls Live (real-time). Multi-page requests aren't a flat charge — each extra page costs 75% of the base rate, so a 10-page deep search adds up to about 7.75× the single-page price. Read from their own pricing page on 10 September 2026.

Serpent's Web category supports up to 10 pages per request. /api/search/quick is billed one published web-search unit per request whatever your tier or balance. /api/search is billed one published web-search unit per page requested, up to 10 pages, at every tier and every balance. Scale tier: $0.03 / 1K pages.

Bright Data's SERP API is $1.50 per 1K requests on pay-as-you-go, dropping to $1.30 per 1K on their Scale plan, which costs $499 a month and includes 380K requests. Their AI Overview extraction is a paid add-on (brd_ai_overview) on top of the base call. Read from their own pricing page on 10 September 2026.

Serpent: AIO is available on eligible opt-in Deep Search calls, with no monthly minimum and $0.03 per 1K at Scale.

Serper.dev's entry tier is $50 for 50K credits ($1 / 1K credits). Their cheapest top-tier rate is $0.30 per 1K credits at high volume. Serper bills 100-result queries as 2 credits and 10-result queries as 1 credit, so a 100-result deep search effectively costs $0.60 per 1K queries at top tier. Serper does not publish a paid ladder on its own site: its /pricing page returned 404 when we checked on 10 September 2026, so those pack rates come from ColdIQ's July 2026 analysis.

Serpent's Growth tier (with a one-time $100 deposit) is $0.06 per 1K pages, and Scale (with a one-time $500 deposit) is $0.03 per 1K pages: Deep Search bills one unit per page requested, while Quick Search is one unit per call whatever depth you ask for. Serpent also offers Google AI Overview fields on eligible opt-in Deep Search calls.

Three reasons:

1. No subscription overhead. Most competitors price for monthly seats and SLAs; we don't carry that fixed cost.
2. Volume discount built in. One-time deposit locks the discount forever — no subscription, no recurring fee.
3. Lean infrastructure. One stack, one architecture, one deployment. We pass those efficiencies through.

The economics work because Serpent is a focused product, not a 20-product enterprise data platform.

Pick Serpent if you want pay-as-you-go billing, the lowest per-call price, optional AI Overview fields on eligible Deep Search calls, and the same JSON shape across Google/Bing/Yahoo/DuckDuckGo. Most SaaS, agency, and rank-tracker workloads fit here.

Pick a different provider if you specifically need Google Scholar or you require an enterprise SLA with 24/7 phone support and contract guarantees.

Send your API key in the X-API-Key header on every request. Generate keys from your dashboard. Keys can be revoked or rotated at any time.

Example: curl -H "X-API-Key: sk_live_..." "https://apiserpent.com/api/search?q=mesothelioma&engine=google&country=us"

Clean JSON. Same top-level shape across all five engines, so you can switch engines by changing one query parameter without rewriting your integration. See the API Reference for the complete schema.

Rate limits are per account (shared across all your API keys) and scale with your live credit balance. They apply to every endpoint, including Quick Search (/api/search/quick) — one shared set of ceilings for the whole account, not a separate quota per endpoint. Each account has four ceilings: max concurrent requests, and max requests per minute, per hour and per day.

Published figures are standard "up to" allocations, not guaranteed floors — effective throughput can be lower under heavy platform load. Each range shows the standard allocation at your balance to the high end available on request: ^* higher custom allocations are provided on request — email info@apiserpent.com for custom limits. For Default accounts, requests above 1,000 calls/day are served on a global shared pool (best effort, not a guaranteed floor). Minute/hour/day windows reset on fixed UTC boundaries; a 429 tells you which window you hit — back off and retry. Your bracket tracks your current balance, so topping up raises limits instantly. Pricing tiers are a separate dimension: they're locked in by a one-time deposit and don't move with your balance.

Pricing and rate limits are two separate dimensions. Your pricing tier is locked by your largest single deposit and never downgrades, but your rate-limit bracket follows your current spendable balance instead. At a balance under $100 you're on the Default bracket (20 requests/minute); at $100–$499 you're on Growth (100/minute); at $500+ you're on Scale (500/minute). Topping up to $100 or $500 raises your bracket instantly, and higher custom allocations are available on request — email info@apiserpent.com for custom limits.

Response time varies by endpoint, engine, result depth, and target country. The Playground shows live timing for your request, and API responses return structured JSON when results are delivered.

Yes. It's a REST API — call it from Python, Node.js, Ruby, Go, PHP, Java, .NET, cURL, n8n, Zapier, Make, anywhere that can make an HTTPS request. The docs include copy-paste examples for the most common stacks.

Search data changes frequently. Serpent API provides a documented JSON contract so your integration can focus on the returned fields and normal API error handling.

Serpent gives your application a stable HTTPS interface and normalized JSON so you can focus on using the data rather than maintaining collection infrastructure.

Yes — try the live Playground from the menu (no signup needed for the demo key). For API integration, sign up to get free calls and run real queries from your code.

Public SERP data is broadly considered legal to scrape under U.S. law (see hiQ Labs v. LinkedIn, Van Buren v. United States) when you're collecting publicly accessible information without authentication or paywall bypass. SERP APIs like Serpent operate within this scope.

That said, "legal to scrape" and "permitted by Terms of Service" are different things — Google's ToS does technically restrict automated access. Most rank-tracking, SEO, and AI agent workloads operate in this practical gray zone, which is why dedicated SERP API providers exist.

Always consult your own legal counsel for your specific use case and jurisdiction.

We log request metadata (your API key, timestamp, endpoint, status code) and the first 200 characters of your search query for billing and abuse prevention, and those logs are deleted on a 90-day cycle. We do not retain the SERP results we return to you — they're fetched for your request, sent to you, and discarded. Bulk and asynchronous jobs are the exception: their results are stored for 7 days so you can download them, then deleted. See our Privacy Policy for full details.

No public SLA is currently published for pay-as-you-go accounts. Enterprise customers can contact info@apiserpent.com to discuss availability requirements.

Every response we could not complete is refunded automatically. One exception: a 404 for a resource that genuinely does not exist is an answer, and that call is charged.

You get an HTTP 402 Payment Required response with { "error": "insufficient_credits" }. Credits are checked atomically per request — there's no overdraft. Top up from your dashboard and resume; nothing else changes.

Serpent API is a product of Contentladder Marketing Pvt Ltd. We've been building search and SEO infrastructure since 2022. Read more on our About page.