AI Summary

AI Summary

curl --request POST \
  --url https://api.cludo.com/api/v4/{customerId}/{engineId}/search/summarize \
  --header 'Authorization: <api-key>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "query": "<string>",
  "queryId": "<string>",
  "sources": [
    {
      "id": "https://help.cludo.com/...",
      "fields": [
        "Title",
        "Content"
      ]
    }
  ],
  "language": "<string>",
  "customPrompt": "<string>",
  "includeCitations": true,
  "tags": [
    "<string>"
  ]
}
'

import requests

url = "https://api.cludo.com/api/v4/{customerId}/{engineId}/search/summarize"

payload = {
    "query": "<string>",
    "queryId": "<string>",
    "sources": [
        {
            "id": "https://help.cludo.com/...",
            "fields": ["Title", "Content"]
        }
    ],
    "language": "<string>",
    "customPrompt": "<string>",
    "includeCitations": True,
    "tags": ["<string>"]
}
headers = {
    "Authorization": "<api-key>",
    "Content-Type": "application/json"
}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {Authorization: '<api-key>', 'Content-Type': 'application/json'},
  body: JSON.stringify({
    query: '<string>',
    queryId: '<string>',
    sources: [{id: 'https://help.cludo.com/...', fields: ['Title', 'Content']}],
    language: '<string>',
    customPrompt: '<string>',
    includeCitations: true,
    tags: ['<string>']
  })
};

fetch('https://api.cludo.com/api/v4/{customerId}/{engineId}/search/summarize', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

const options = {
  method: 'POST',
  headers: {Authorization: '<api-key>', 'Content-Type': 'application/json'},
  body: JSON.stringify({
    query: '<string>',
    queryId: '<string>',
    sources: [{id: 'https://help.cludo.com/...', fields: ['Title', 'Content']}],
    language: '<string>',
    customPrompt: '<string>',
    includeCitations: true,
    tags: ['<string>']
  })
};

fetch('https://api.cludo.com/api/v4/{customerId}/{engineId}/search/summarize', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

{
  "value": {
    "summary": "<string>",
    "summaryRequestId": "<string>",
    "summaryId": "<string>",
    "summarySourceUrls": [
      "<string>"
    ]
  }
}

{
  "message": "The request could not be processed. Check the request body and required parameters."
}

{
  "message": "unauthorized"
}

{
  "message": "The provided sources did not contain enough relevant context to generate a summary."
}

{
  "status": "InternalServerError",
  "code": 500,
  "messages": [
    "Error occur. Please contact support with error ID: a1b2c3d4-e5f6-7890-abcd-ef1234567890"
  ]
}

POST

api

{customerId}

{engineId}

summarize

AI Summary

curl --request POST \
  --url https://api.cludo.com/api/v4/{customerId}/{engineId}/search/summarize \
  --header 'Authorization: <api-key>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "query": "<string>",
  "queryId": "<string>",
  "sources": [
    {
      "id": "https://help.cludo.com/...",
      "fields": [
        "Title",
        "Content"
      ]
    }
  ],
  "language": "<string>",
  "customPrompt": "<string>",
  "includeCitations": true,
  "tags": [
    "<string>"
  ]
}
'

import requests

url = "https://api.cludo.com/api/v4/{customerId}/{engineId}/search/summarize"

payload = {
    "query": "<string>",
    "queryId": "<string>",
    "sources": [
        {
            "id": "https://help.cludo.com/...",
            "fields": ["Title", "Content"]
        }
    ],
    "language": "<string>",
    "customPrompt": "<string>",
    "includeCitations": True,
    "tags": ["<string>"]
}
headers = {
    "Authorization": "<api-key>",
    "Content-Type": "application/json"
}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {Authorization: '<api-key>', 'Content-Type': 'application/json'},
  body: JSON.stringify({
    query: '<string>',
    queryId: '<string>',
    sources: [{id: 'https://help.cludo.com/...', fields: ['Title', 'Content']}],
    language: '<string>',
    customPrompt: '<string>',
    includeCitations: true,
    tags: ['<string>']
  })
};

fetch('https://api.cludo.com/api/v4/{customerId}/{engineId}/search/summarize', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

const options = {
  method: 'POST',
  headers: {Authorization: '<api-key>', 'Content-Type': 'application/json'},
  body: JSON.stringify({
    query: '<string>',
    queryId: '<string>',
    sources: [{id: 'https://help.cludo.com/...', fields: ['Title', 'Content']}],
    language: '<string>',
    customPrompt: '<string>',
    includeCitations: true,
    tags: ['<string>']
  })
};

fetch('https://api.cludo.com/api/v4/{customerId}/{engineId}/search/summarize', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

{
  "value": {
    "summary": "<string>",
    "summaryRequestId": "<string>",
    "summaryId": "<string>",
    "summarySourceUrls": [
      "<string>"
    ]
  }
}

{
  "message": "The request could not be processed. Check the request body and required parameters."
}

{
  "message": "unauthorized"
}

{
  "message": "The provided sources did not contain enough relevant context to generate a summary."
}

{
  "status": "InternalServerError",
  "code": 500,
  "messages": [
    "Error occur. Please contact support with error ID: a1b2c3d4-e5f6-7890-abcd-ef1234567890"
  ]
}

Standard
Streaming

You are viewing the standard endpoint which returns a complete JSON response. Generate AI-powered summaries from search result documents. The AI Summary endpoint works in tandem with the Search endpoint: first execute a search, then pass the search results as sources to generate a concise summary with citations.

The streaming endpoint is the same path as standard with /stream appended: POST /api/v4/{customerId}/{engineId}/search/summarize/stream. It accepts the same request body as the standard endpoint, but streams the response as chunked plain text. Streaming responses set Cludo-Summary-Id and Cludo-Summary-Request-Id headers for tracking.

Workflow

Step 1: Execute a search Call the Search endpoint and extract from the response:

QueryId → use as the queryId parameter
TypedDocuments[].Fields.Id.Value → use as the id for each source
Source fields → prefer Title and Content when those fields are available

Step 2: Generate the summary Pass the extracted values to this endpoint along with your query and source configuration. Send no more than 5 sources.

Feature availability

answerLength and customPrompt are in limited availability. These options are currently in an early test phase and are only enabled for select customers running on specific servers. They are not active by default, so supplying these values will not change the generated output unless the features have been enabled for your account.To discuss enabling them for your setup, contact your account manager and share your customer ID and engine ID.

includeCitations controls inline citation links in the summary text. When true, the model adds numbered Markdown links (e.g. [1](https://example.com/page)) directly inside summary for each cited source. When false, the summary is plain text with no inline links. This applies to both the standard and streaming endpoints. summarySourceUrls is always returned regardless of this flag.

Tagging requests for analytics

The optional tags field labels the context a summary was requested from. Tags are stored on the analytics events for the request and do not change the summary. They work on both the standard and streaming endpoints. If you built AI Mode yourself, see Tagging AI Mode requests for the tag that makes its requests count as AI Mode.

{
  "query": "return policy",
  "queryId": "<QueryId from the search response>",
  "sources": [
    { "id": "https://www.example.com/returns", "fields": ["Title", "Content"] }
  ],
  "tags": ["<string>"]
}

A request is never rejected because of its tags. Instead, the tags are sanitized:

Each tag is lower-cased, and characters other than a–z, 0–9 and _ are removed. This also removes surrounding whitespace.
Each tag is cut to 32 characters.
Empty tags, duplicate tags and entries that are not strings are dropped.
Only the first 5 tags are kept.
A tags value that is not an array is ignored.

Authorizations

Authorization

string

header

required

SiteKey authentication. Pass the full value including the SiteKey prefix: SiteKey <base64(customerId:engineId:searchKey)>. Example: SiteKey dGVzdDoxMjM0NTY3Ojk4NzY1.

Path Parameters

customerId

integer

required

Your Cludo account ID.

engineId

integer

required

Search engine ID.

Body

application/json

query

string

required

Original user query string from search (for context and analytics).

queryId

string

required

Must equal QueryId from the search response that produced sources.

sources

object[]

required

One entry per result row to fold into the summary (typically the checked results on the SERP). Include at least three sources and no more than five.

Maximum array length: 5

Show child attributes

Example:

[
  {
    "id": "https://help.cludo.com/...",
    "fields": ["Title", "Content"]
  }
]

language

string

ISO-639-1 or simple locale code for answer language (e.g. en, da).

answerLength

enum<string>

comprehensive for longer explanations; concise for shorter summaries. Limited availability: this feature is currently in an early test phase and is only enabled for select customers on specific servers. It is not active by default, so the value has no effect unless it has been enabled for your account. Contact your account manager (with your customer ID and engine ID) to discuss enabling it.

Available options:

comprehensive,

concise

customPrompt

string

Optional extra instructions (e.g. “focus on steps”). Limited availability: this feature is currently in an early test phase and is only enabled for select customers on specific servers. It is not active by default, so the value has no effect unless it has been enabled for your account. Contact your account manager (with your customer ID and engine ID) to discuss enabling it.

includeCitations

boolean

When true, the summary includes inline Markdown citation links (e.g. [1](url)) for each cited source, directly inside the summary text. When false (default), the summary is plain text with no inline links. Applies to both the standard and streaming endpoints. summarySourceUrls is returned either way, independent of this flag.

Response

AI Summary generated successfully.

value

object

Show child attributes

Overview

Search

AI

Content & Indexing

Tracking

Feedback

Search Tools

Workflow

Feature availability

Tagging requests for analytics

Authorizations

Path Parameters

Body

Response

​Workflow

​Feature availability

​Tagging requests for analytics

Authorizations

Path Parameters

Body

Response

Workflow

Feature availability

Tagging requests for analytics