2026, September 30
AI Overview extraction
AI Overview extraction
Retrieve Google's AI-generated answer box directly from the Google Search endpoint. Add ai_overview=1 to any Google search request to receive the full AI Overview text, structured paragraphs, and cited sources when Google renders one for that query.
- New parameter: ai_overview=1 on GET /api/v2/search
- Response includes ai_overview.present, text, text_paragraphs[], and citations[] (title, url, domain, source, position)
- When no AI Overview is rendered for a query, the response returns ai_overview.present: false and falls back to a standard search result — no error, no nulls
- Billed at 3× the standard credit rate, applied only on successful responses where present: true
- Absent AIO (present: false) billed at the standard 1× rate
- Available on: Small, Medium, Large, Premium, Enterprise
Answer Engines (ChatGPT, Perplexity, Gemini, Copilot)
Query generative AI search engines via a single API endpoint. Send a prompt and receive the engine's answer together with all cited sources. Available engines depend on plan tier.
- New endpoint: GET /api/v2/answer
- Parameters: engine, prompt, additional_prompt (optional), country (optional), brand (optional)
- Response includes answer_text, model, web_search_triggered, citations[], search_sources[]
- Optional brand parameter returns a brand_visibility object showing whether the brand is cited and at what position
- Billed at 7× the standard credit rate per successful answer, flat rate across all engines
- Engine availability by plan:
- Medium: ChatGPT
- Large: + Perplexity
- Premium: + Gemini
- Enterprise: + Copilot (all four engines)
- Available on: Medium, Large, Premium, Enterprise
Markdown output
All search endpoints now support output=markdown, returning clean, LLM-ready Markdown instead of JSON. Designed for AI agent and RAG pipeline use cases.
- New parameter: output=markdown (alongside existing json and html) on all search endpoints
- Strips navigation, ads, and boilerplate — returns readable Markdown suitable for direct use as LLM context
- No additional cost — billed at the standard per-request rate
- Available on: Small, Medium, Large, Premium, Enterprise
Dedicated verticals — Hotels, Flights, Jobs
Three new dedicated SERP verticals, each with structured, vertical-specific parameters and parsed JSON responses.
Hotels Search — tbm=hotels
- Parameters: q, check_in, check_out, adults, children, currency, free_cancellation
- Returns hotel listings with pricing, ratings, availability, and amenities
Flights Search — tbm=flights
- Parameters: departure_id, arrival_id, outbound_date, return_date, currency
- Returns itineraries with per-leg details (airline, flight number, departure, arrival, duration, stops)
Jobs Search — tbm=jobs
- Parameters: q, location, num, start
- Returns job listings with title, company, location, salary, schedule, and full description
All three verticals are available on: Small, Medium, Large, Premium, Enterprise
2026, April 1
The num parameter has been removed from the Zenserp API to simplify request handling and align with current response behavior; all documentation, examples, and code snippets have been updated, and users should remove this parameter from their integrations to ensure compatibility.
Removed the
numparameter from the Zenserp API documentation and OpenAPI specificationUpdated all examples and code snippets to reflect this change
Synced documentation with the latest API specification updates