Up to 100 results per call, with AI Overview support.
Pages 2 to 10 priced lower than page 1.
Send this request
1curl https://api.zyte.com/v1/search \
2 --user YOUR_ZYTE_API_KEY: \
3 --header 'Content-Type: application/json' \
4 --data '{
5 "domain": "search.engine.com",
6 "query": "running shoes",
7 "include": ["organic", "aiOverview"],
8 "maxResults": 100
9 }'Get this response
1{
2 "status": "success",
3 "url": "https://www.search.engine.com/search?q=running+shoes",
4 "fetchedAt": "2026-05-21T09:36:57Z",
5 "organicResults": [
6 {
7 "rank": 1,
8 "title": "Best Running Shoes 2026",
9 "url": "https://example.com/best-running-shoes",
10 "snippet": "After testing 40+ pairs..."
11 },
12 {
13 "rank": 2,
14 "title": "Top 10 Running Shoes for Marathoners",
15 "url": "https://example.org/marathon-shoes",
16 "snippet": "Expert rankings for every distance..."
17 },
18 {
19 "rank": 3,
20 "title": "How to Choose the Right Running Shoe",
21 "url": "https://example.net/choose-running-shoes",
22 "snippet": "Find your perfect fit..."
23 }
24 // 100 results, fetched across 10 pages
25 ]
26}
At a few hundred keywords, SERP collection is a project an engineer ships in a sprint. At a few hundred thousand, it's a system three people maintain full-time. By the time it's become infrastructure, switching is harder than building it was.
Most SERP APIs charge a flat per-page rate because they fetch each page as a separate request - new browser session, new proxy, new ban check, every time. Ten pages means ten times the infrastructure cost.
Coordinated pagination is different. One call returns up to 100 results. Page 1 warms the session; pages 2 to 10 ride on the same infrastructure. They cost less to run, so they cost less to buy.
One endpoint, multiple shapes of data. Configure your request to pull organic results, AI Overviews, geo-targeted rankings, or all of them, in a single call.
Capture AI Overview content within the page HTML, returned inside the JSON response - parse it alongside your structured organic results. As LLM-generated answers reshape SERPs, your data stays current.
"include": ["aiOverview"]
Don't paginate. Pages 1 through 10 fetched in the same coordinated workflow - and pages 2-10 cost less.
"maxResults": 100
Target any country, language, or specific city. Different rankings for different markets, in a single parameter.
"queryParameters": { "gl": "us" }
Most SERP APIs charge a flat per-page rate. We charge less for pages 2 through 10 because they share infrastructure with the page-1 session. The deeper you paginate, the better the unit economics get.
| PAYG | $100 / month | $200 / month | $500 / month | |
|---|---|---|---|---|
| First page | $0.85 / 1K responses | $0.63 / 1K responses | $0.51 / 1K responses | $0.40 / 1K responses |
Pages 2 to 10 ~29% less | $0.61 / 1K responses | $0.45 / 1K responses | $0.36 / 1K responses | $0.29 / 1K responses |
A lower headline price usually means lighter infrastructure on pages 2-10. The difference shows up as your QA and re-collection cost, not on the invoice. Compare cost per reliable, complete result.
I love the ease of the initial setup with Zyte, as they took care of all the development, and we only needed to communicate what data we needed and set up the necessary processes on our end. I find it great that Zyte handles everything for us, allowing us not to worry about maintaining a platform ourselves.
Zyte API is built to be resilient to ongoing SERP changes, with continuously updated retrieval and extraction logic so customers don’t need to rebuild their pipelines.
The launched default is tuned for roughly 90%+ success at high volume; higher rates are achievable with configuration. Reliability is maintained through adaptive unblocking infrastructure, including optimized proxy routing, session management, rendering, and resilient request orchestration.
Zyte API supports geo-targeted search collection for localized results, including country, language and regional targeting. Support for highly localized search scenarios depends on the search engine and the level of location precision required.
Zyte API supports coordinated pagination from num=10 up to num=100, desktop search collection, geo-targeting, structured organic results, and HTML responses that include AI-generated search features when present.
Zyte API returns responses as JSON, with the raw HTML included as a field within the JSON response object.
Zyte API supports synchronous requests. Asynchronous requests are planned as part of the roadmap.
Yes. Zyte API uses request per minute (RPM) limits based on your plan.Platform plans support up to 3,000 RPM. Enterprise plans support up to 10,000 RPM, with higher throughput available through custom enterprise agreements. If you need more than 10,000 RPM, contact sales to discuss scaling requirements.
Zyte API uses success-based billing. You are charged only for successful requests returned by the API, with flat pricing per successful request based on your plan.
Yes. Usage can vary depending on factors such as pagination depth, rendering requirements, extraction type, and other workflow characteristics
No. Unused monthly usage does not roll over. However, Zyte API plans support overages. For example, if you commit to a $100/month plan, you can continue using the platform up to $200/month before needing to increase your plan limit.
Yes. Zyte offers a free trial with a $5 credit that allows users to test the API and SERP capabilities before committing to a paid plan - start free. If you want to test at scale, talk to us about setting you up on an Enterprise Trial with a $100 credit.
Yes. Zyte offers volume-based pricing and additional discounts for Enterprise customers with larger scale or long-term usage requirements.
SERP data is structured information from search results pages, including rankings, URLs, titles, snippets, and rich result features. Teams use it for rank tracking, competitive intelligence, SEO analysis, and feeding AI systems with search-derived data.
Most SERP APIs treat each page as a separate request. Zyte treats SERP collection as a workflow - coordinating pagination, retries, and data extraction across multiple pages in a single execution. This improves reliability and reduces cost at scale.
Workflow-based extraction means multiple SERP pages are retrieved as part of a coordinated process rather than isolated requests. Zyte optimizes execution across pages - handling session reuse, retries, and infrastructure selection automatically.
Zyte reduces cost by optimizing across the full SERP workflow instead of pricing each page independently. Features like num=100 workflows, session reuse, and adaptive infrastructure lower the total number of high-cost requests required per keyword.
num=100 refers to retrieving up to 100 search results (multiple pages) in a single workflow. Instead of issuing separate requests per page, Zyte executes them as one coordinated process, significantly reducing cost and improving efficiency for deep pagination use cases.
Zyte pricing is based on successful pages returned, not attempted requests. If some pages fail, you are only billed for the pages successfully delivered. This aligns cost with outcomes and helps maintain predictable unit economics.
Yes. Zyte API is designed for high-volume SERP workloads, including millions of keywords. The platform supports burst traffic, deep pagination, and large datasets while maintaining consistent performance and predictable costs.
Zyte uses adaptive execution strategies, including proxy selection, session management, and retry logic, to maintain high success rates. This ensures stable data pipelines even as target sites change or introduce blocking mechanisms.
Zyte supports extraction of standard organic results as well as evolving SERP components, such as structured result blocks and AI-generated features when present. Data is returned in a structured format ready for analysis or ingestion.
SERP data can be delivered in structured formats such as JSON. It can also be integrated into your data pipelines with no additional processing required.