Request & Response - Webz.io

Endpoint

POST

https://api.webz.io/api/news/answers

Authentication

Use your Webz API token in one of these ways:

Recommended: send Authorization: Bearer YOUR_API_KEY.
Alternative: add ?token=YOUR_API_KEY to the endpoint URL.

Your account must have the api_news permission and enough credits or PAYG balance for the selected answer type.

Request body

Send both parameters in a JSON body.

Parameter	Type	Required	Default	Description
`query`	string	Yes	—	Natural-language question. Maximum 750 characters and 100 words.
`detailed`	boolean	No	`false`	Set to `true` for a longer answer based on more evidence.

The API infers the searches, filters, time window, language, and sort order from query. You do not send News Search filters or result-count parameters.

Standard answer

curl -s -X POST "https://api.webz.io/api/news/answers" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "query": "Why has Nvidia been under regulatory scrutiny in the past week?",
    "detailed": false
  }'

Detailed answer

curl -s -X POST "https://api.webz.io/api/news/answers" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "query": "Why has Nvidia been under regulatory scrutiny in the past week?",
    "detailed": true
  }'

Response

A successful request returns HTTP 200.

json

{
  "query": "Why has Nvidia been under regulatory scrutiny in the past week?",
  "detailed": true,
  "answer": {
    "text": "Regulators have focused on Nvidia's market position and business practices [1].",
    "citations": [
      {
        "id": "article-uuid_0",
        "title": "Regulators examine Nvidia's market position",
        "url": "https://news.example.com/nvidia-regulatory-review",
        "published": "2026-10-18T09:30:00Z",
        "supporting_text": "Regulators are examining Nvidia's market position and business practices."
      }
    ],
    "insufficient_evidence": false
  },
  "total_results": 1,
  "results": [
    {
      "score": 8.7,
      "article": {
        "article_id": "article-uuid",
        "url": "https://news.example.com/nvidia-regulatory-review",
        "title": "Regulators examine Nvidia's market position",
        "published_at": "2026-10-18T09:30:00Z"
      },
      "chunk": {
        "chunk_id": "article-uuid_0",
        "chunk_index": 0,
        "text": "Regulators are examining Nvidia's market position and business practices."
      },
      "metadata": {
        "language": "english",
        "domain": "news.example.com",
        "site_type": "news"
      }
    }
  ],
  "credits_used": 8,
  "requests_left": 992
}

Field	Description
`query`	The submitted question after surrounding whitespace is removed.
`detailed`	The answer mode used for this request.
`answer.text`	Generated answer. Markers such as `[1]` refer to citations.
`answer.citations`	Sources cited by the answer, in marker order.
`answer.citations[].id`	Supporting chunk ID.
`answer.citations[].title`	Source article title.
`answer.citations[].url`	Source article URL.
`answer.citations[].published`	Source publication time.
`answer.citations[].supporting_text`	Source excerpt supporting the answer.
`answer.insufficient_evidence`	`true` when the available news does not support an answer.
`total_results`	Number of unique supporting chunks returned.
`results`	Supporting News Search results. Cited chunks come first, followed by uncited results.
`credits_used`	Credits charged for a subscription request.
`requests_left`	Remaining subscription credit balance.
`cost`	Amount charged for a PAYG request.
`balance`	PAYG balance after the charge.

Citation [n] corresponds to answer.citations[n - 1] and results[n - 1]. Each citation's id matches the corresponding results[].chunk.chunk_id.

The objects in results use the same article, chunk, and metadata fields documented in News Search Response Format.