Search API

Execute a search queries using Nimble Search

Execute web searches and retrieve clean, parsed content from top results in two modes:

  • Fast Mode — Quickly discovers high-value URLs or generates fast web answers for your agents It’s ideal for lightweight discovery or pairing with /extract when you want to selectively fetch full content only from the URLs you choose. Costs 1 credit per search.

  • Deep Search — Performs real-time webpage extraction for full, rich context, automatically calling /extract to retrieve complete webpage content for deeper research tasks. Deep Search costs 1 credit per search, plus 1 additional credit for each webpage extracted.

Both Fast Mode and Deep Search support advanced search filtering, location targeting, max result limits, and additional web-search refinement tools.

Basic Usage

curl -X POST https://nimble-retriever.webit.live/search \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "latest AI trends 2025",
    "num_results": 10,
    "deep_search": false,
    "country": "US",
    "locale": "en"
  }'

Response

{
  "message": "Request processed successfully",
  "body": [
    {
      "page_content": "",
      "metadata": {
        "title": "Latest GenAI Trends: 5 Key Developments to Watch",
        "snippet": "Explore the top trends in generative AI including agentic AI, retrieval-augmented generation, self-training models, and ethical AI implementations.",
        "url": "https://example.com/genai-trends-2025",
        "position": 1,
        "entity_type": "OrganicResult",
        "country": "US",
        "locale": "en"
      }
    },
    {
      "page_content": "",
      "metadata": {
        "title": "GenAI Investment and Market Analysis 2025",
        "snippet": "Despite challenges, investment in generative AI continues to grow. Analysis of market trends, adoption rates, and future projections for enterprise AI.",
        "url": "https://example.com/genai-investment-analysis",
        "position": 2,
        "entity_type": "OrganicResult",
        "country": "US",
        "locale": "en"
      }
    },
    {
      "page_content": "",
      "metadata": {
        "title": "Emerging Technologies: AI and Data Trends",
        "snippet": "Comprehensive overview of agent-ready data, AI for data engineering, data provenance, compliance changes, and the rise of agentic edge computing.",
        "url": "https://example.com/emerging-tech-trends",
        "position": 3,
        "entity_type": "OrganicResult",
        "country": "US",
        "locale": "en"
      }
    }
  ]
}

Fast Mode with - Include Answer

Optional AI-generated answer summaries that provide quick insights from search results without reading full content. This costs 1 additional credit per request.

Response:

Response


Request Body Parameters

Parameter
Type
Required
Default
Description

query

string

Yes

-

The search query to execute

num_results

integer

No

3

Number of search results to return (max: 100)

topic

enum["general", "news", "location"]

No

general

The search topic type. news returns real-time news articles with publication dates. location returns places with ratings, addresses, and reviews (restaurants, businesses, venues). general (default) returns standard web search results across all sources.

deep_search

boolean

No

true

When true, fetches and parses full page content. When false, return only meta_data.

parsing_type

enum

No

plain_text

Output format. Options: plain_text, markdown, simplified_html

locale

string

No

en

Locale for search results (e.g., en, fr, de)

country

string

No

US

Country code for search results (e.g., US, FR, GB)

include_answer

boolean

No

false

Generate LLM answer summary (only available when deep_search=False)

include_domains

string[]

No

List of domains to include in search results. Maximum 50 domains.

exclude_domains

string[]

No

List of domains to exclude from search results. Maximum 50 domains.

start_date

string

No

Filter results after this date (format: YYYY-MM-DD or YYYY)

end_date

string

No

Filter results before this date (format: YYYY-MM-DD or YYYY)


Last updated