For AI agents: markdown of this page — /docs-content-en/calls/followup.md documentation index — /llms.txt

Call follow-up

⚠️ The methods ship in the call 26.600.0 update and are not yet available on all Bitrix24 accounts. If the update has not reached your Bitrix24 account, the API returns 422 METHOD_NOT_YET_AVAILABLE — this means the method has not been released on the account yet, not an integration error.

Read the AI Follow-up of completed calls: transcription, meeting overview, summary, insights, and efficiency evaluation. The methods only read data and change nothing.

Scope: call | Base URL: https://vibecode.bitrix24.com/v1 | Authorization: X-Api-Key

Follow-up list

⚠️ The method ships in update call 26.600.0 and is not yet available on all Bitrix24 accounts. If the update has not reached your account yet, the API returns 422 METHOD_NOT_YET_AVAILABLE — this means the method has not been released on the account, not that the integration is broken.

POST /v1/calls/followups/list

Returns a list of AI Follow-ups for completed calls within the specified period. Pagination through the results is cursor-based.

Request fields (body)

Field	Type	Req.	Description
`filter`	object	yes	Filter conditions. Without this object the request is rejected
`filter.startDate.from`	string	yes	Start of the period, ISO 8601 — `2026-01-01T00:00:00Z`
`filter.startDate.to`	string	yes	End of the period, ISO 8601. A value earlier than `from` is rejected
`filter.participantId`	integer	no	Filter by a single participant. Available to the Bitrix24 account administrator only: a regular employee cannot filter by another participant. List of employees — `GET /v1/users`
`select`	string[]	no	Which fields to return. Allowed values — in the "Select field" section
`order`	object	no	Sorting by start date: `{ "startDate": "asc" }` or `{ "startDate": "desc" }`. Defaults to `desc`
`pagination`	object	no	Page size and cursor. Rules — in the "Navigation" section
`pagination.limit`	integer	no	Records per page. Defaults to 50, maximum 200. Selecting data-heavy fields lowers the maximum to 20
`pagination.afterCursor`	object	no	Cursor of the next page. Copied as a whole from `data.afterCursor` of the previous response. Not passed in the first request
`mentionFormat`	string	no	Format of `@`-mentions in text fields: `bb`, `html` or `none`. Defaults to `bb`

Select field

Without select only call metadata is returned. The remaining fields are requested explicitly. An unknown value is rejected with 422 BITRIX_ERROR and a message that names the rejected field.

Metadata

Value	What is returned
`callId`, `callType`, `initiatorId`, `startDate`, `endDate`, `durationSeconds`	Base fields. Always returned in the list
`uuid`	Call session identifier
`participants`	List of participants
`tracks`	Call recordings with download links
`outcomes`	List of the AI blocks that are ready
`language`	Transcription language
`version`	Version of the AI data schema
`createdAt`	Time of the last AI record

AI blocks

A block is requested either as a whole or as an individual subfield using dot notation.

Block	Subfields
`transcription`	`transcription.language`, `transcription.segments`
`overview`	`overview.topic`, `overview.detailedTakeaways`, `overview.meetingType`, `overview.agenda`, `overview.agreements`, `overview.actionItems`, `overview.meetings`
`summary`	Requested only as a whole
`insights`	`insights.speakerEvaluationAvailable`, `insights.speakerAnalysis`, `insights.meetingStrengths`, `insights.meetingWeaknesses`, `insights.speechStyleInfluence`, `insights.engagementLevel`, `insights.areasOfResponsibility`, `insights.finalRecommendations`
`evaluation`	`evaluation.efficiencyValue`, `evaluation.calendar`, `evaluation.criteria`

Selection rules:

Requesting a whole block (["overview"]) returns all of its subfields.
Requesting a subfield (["overview.topic"]) returns only that subfield — the rest are omitted.
transcription, transcription.segments, overview and insights contain a lot of data. Selecting them lowers the maximum page size to 20 records.

The first request is sent without pagination.afterCursor.
If hasMore in the response is true, copy the whole data.afterCursor object into pagination.afterCursor of the next request.
Repeat until hasMore becomes false.

pagination.limit — defaults to 50, maximum 200. Selecting data-heavy fields lowers the maximum to 20.

jsonc

{ "pagination": { "limit": 50, "afterCursor": { "startDate": "2026-01-12T14:30:00.000000+00:00", "id": 12330 } } }

Mention format

mentionFormat controls how @-mentions of employees look in all AI text fields — transcription.segments[].text, overview.*, insights.*, evaluation.criteria.*.thoughts.

Value	Text in the response	When to choose
`bb`	`[USER=7]John Smith[/USER]`	Rendering in an interface that supports BB code
`html`	`<span class="bx-call-mention" …>John Smith</span>`	Rendering on the web
`none`	`John Smith`	Passing the text to an AI model, search, export

Examples

A review of one employee's meetings for January: we take the call metadata, the topic, the agreements and the action items from the overview, plus the participant analysis and the overall score. Results are sorted newest first, and mentions arrive as plain text. This is the first request, so it has no afterCursor — the cursor for the next page arrives in the response.

curl — personal key

Terminal

curl -X POST https://vibecode.bitrix24.com/v1/calls/followups/list \
  -H "X-Api-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "filter": {
      "startDate": { "from": "2026-01-01T00:00:00Z", "to": "2026-01-31T23:59:59Z" },
      "participantId": 42
    },
    "select": [
      "callId", "startDate", "durationSeconds", "participants", "outcomes",
      "overview.topic", "overview.agreements", "overview.actionItems",
      "insights.speakerAnalysis", "evaluation.efficiencyValue"
    ],
    "order": { "startDate": "desc" },
    "pagination": { "limit": 20 },
    "mentionFormat": "none"
  }'

curl — OAuth application

Terminal

curl -X POST https://vibecode.bitrix24.com/v1/calls/followups/list \
  -H "X-Api-Key: YOUR_APP_KEY" \
  -H "Authorization: Bearer USER_SESSION_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "filter": {
      "startDate": { "from": "2026-01-01T00:00:00Z", "to": "2026-01-31T23:59:59Z" },
      "participantId": 42
    },
    "select": [
      "callId", "startDate", "durationSeconds", "participants", "outcomes",
      "overview.topic", "overview.agreements", "overview.actionItems",
      "insights.speakerAnalysis", "evaluation.efficiencyValue"
    ],
    "order": { "startDate": "desc" },
    "pagination": { "limit": 20 },
    "mentionFormat": "none"
  }'

JavaScript — personal key

javascript

const res = await fetch('https://vibecode.bitrix24.com/v1/calls/followups/list', {
  method: 'POST',
  headers: {
    'X-Api-Key': 'YOUR_API_KEY',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    filter: {
      startDate: { from: '2026-01-01T00:00:00Z', to: '2026-01-31T23:59:59Z' },
      participantId: 42,
    },
    select: [
      'callId', 'startDate', 'durationSeconds', 'participants', 'outcomes',
      'overview.topic', 'overview.agreements', 'overview.actionItems',
      'insights.speakerAnalysis', 'evaluation.efficiencyValue',
    ],
    order: { startDate: 'desc' },
    pagination: { limit: 20 },
    mentionFormat: 'none',
  }),
})

const body = await res.json()

if (!body.success) {
  // Handle the error explicitly: otherwise on 403 or 422 the code fails when accessing data
  throw new Error(`${body.error.code}: ${body.error.message}`)
}

const { data } = body

for (const call of data.items) {
  // AI blocks are not ready for every call: some of them arrive as null
  const topic = call.overview?.topic ?? 'topic not determined'
  const score = call.evaluation?.efficiencyValue ?? '—'
  console.log(call.startDate, topic, `score: ${score}`)

  for (const item of call.overview?.actionItems ?? []) {
    console.log('  action item:', item.actionItem)
  }
}

// Next page: put the cursor from the response into pagination.afterCursor of the new request
console.log(data.hasMore, data.afterCursor)

JavaScript — OAuth application

javascript

const res = await fetch('https://vibecode.bitrix24.com/v1/calls/followups/list', {
  method: 'POST',
  headers: {
    'X-Api-Key': 'YOUR_APP_KEY',
    'Authorization': 'Bearer USER_SESSION_TOKEN',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    filter: {
      startDate: { from: '2026-01-01T00:00:00Z', to: '2026-01-31T23:59:59Z' },
      participantId: 42,
    },
    select: [
      'callId', 'startDate', 'durationSeconds', 'participants', 'outcomes',
      'overview.topic', 'overview.agreements', 'overview.actionItems',
      'insights.speakerAnalysis', 'evaluation.efficiencyValue',
    ],
    order: { startDate: 'desc' },
    pagination: { limit: 20 },
    mentionFormat: 'none',
  }),
})

const body = await res.json()

if (!body.success) {
  // Handle the error explicitly: otherwise on 403 or 422 the code fails when accessing data
  throw new Error(`${body.error.code}: ${body.error.message}`)
}

const { data } = body
console.log(data.items, data.hasMore, data.afterCursor)

Response fields

Field	Type	Description
`success`	boolean	Always `true` on success
`data.items`	object[]	Array of Follow-ups. Element fields — in the table below
`data.hasMore`	boolean	Whether there are records beyond the current page
`data.afterCursor`	object \| null	Cursor of the next page. `null` — there are no more records

Fields of an array element:

Field	Type	Description
`items[].callId`	number	Call identifier. Passed to `GET /v1/calls/followups/:callId`
`items[].callType`	number	Call type: `1` — instant, `2` — conference, `3` — big room
`items[].initiatorId`	number	Call initiator. Employee profile — `GET /v1/users/:id`
`items[].startDate`	string	Call start, ISO 8601 in UTC
`items[].endDate`	string \| null	Call end, ISO 8601 in UTC
`items[].durationSeconds`	number	Call duration in seconds
`items[].uuid`	string	Call session identifier
`items[].language`	string	Transcription language code
`items[].version`	number	Version of the AI data schema
`items[].participants`	object[]	Call participants
`items[].outcomes`	string[]	AI blocks already generated for this call
`items[].createdAt`	string	Time of the last AI record
`items[].tracks`	object[]	Call recordings
`items[].transcription`	object	Transcription of the conversation
`items[].overview`	object	Meeting overview
`items[].summary`	object	Fragment-by-fragment summary
`items[].evaluation`	object	Meeting efficiency score
`items[].insights`	object	Insights and participant analysis

Participants — `participants[]`

Field	Type	Description
`userId`	number	Employee identifier
`name`	string	Employee name
`avatar`	string	Link to the avatar
`workPosition`	string	Job title
`talkedSeconds`	number	How many seconds the participant spoke

Call recordings — `tracks[]`

Field	Type	Description
`trackId`	number	Recording identifier
`type`	string	Track type, for example `mixed_audio`
`duration`	number	Recording duration in seconds
`fileName`	string	File name
`mimeType`	string	MIME type of the file
`url`	string	Download link
`dateCreate`	string	When the recording was created, ISO 8601

Transcription — `transcription`

Field	Type	Description
`language`	string	Transcript language code
`segments[]`	object[]	Utterances in chronological order
`segments[].userId`	number	Who is speaking
`segments[].userName`	string	Speaker name
`segments[].start`	number	Start of the utterance, seconds from the start of the call
`segments[].end`	number	End of the utterance, seconds from the start of the call
`segments[].text`	string	Text of the utterance

Meeting overview — `overview`

Field	Type	Description
`topic`	string	Meeting topic
`detailedTakeaways`	string	Detailed takeaways
`meetingType`	object	Meeting type: `typeTag` — code, `title` — name, `explanation` — justification
`agenda`	object	Agenda: `explanation` — wording, `quote` — quote from the conversation
`agreements[]`	object[]	Agreements: `agreement` — wording, `quote` — quote
`actionItems[]`	object[]	Action items: `actionItem` — wording, `quote` — quote
`meetings[]`	object[]	Scheduled meetings: `meeting` — wording, `quote` — quote

Summary — `summary`

Field	Type	Description
`segments[]`	object[]	Fragments of the conversation
`segments[].start`	number	Start of the fragment, seconds
`segments[].end`	number	End of the fragment, seconds
`segments[].title`	string	Fragment title
`segments[].summary`	string	Summary of the fragment

Insights — `insights`

Field	Type	Description
`speakerEvaluationAvailable`	boolean	Whether participant analysis is available on the Bitrix24 account
`speakerAnalysis[]`	object[]	Participant analysis, in descending order of `talkPercentage`
`speakerAnalysis[].userId`	number	Employee identifier. Profile — `GET /v1/users/:id`
`speakerAnalysis[].detailedInsight`	string	Conclusion about the participant
`speakerAnalysis[].efficiencyValue`	number	Participant efficiency score, from 0 to 100
`speakerAnalysis[].evaluationCriteria`	string	The criterion the score was based on
`speakerAnalysis[].talkPercentage`	number	Share of the time the participant spoke, in percent
`speakerAnalysis[].duration`	number	How many seconds the participant spoke
`speakerAnalysis[].durationFormat`	string	The same time in `MM:SS` format
`meetingStrengths[]`	object[]	Strengths: `strengthTitle` and `strengthExplanation`
`meetingWeaknesses[]`	object[]	Weaknesses: `weaknessTitle` and `weaknessExplanation`
`speechStyleInfluence`	string	How the speaking style influenced the meeting
`engagementLevel`	string	Engagement of the participants
`areasOfResponsibility`	string	Who is responsible for what after the meeting
`finalRecommendations`	string	Recommendations for the next meeting

Efficiency score — `evaluation`

Field	Type	Description
`efficiencyValue`	number	Overall meeting efficiency, from 0 to 100
`calendar.overhead`	boolean	Whether the meeting ran longer than necessary
`criteria`	object	Map of criteria. The key is the criterion code, the value is an object with the fields below
`criteria.<code>.value`	boolean	Whether the criterion is met
`criteria.<code>.title`	string	Criterion name
`criteria.<code>.criteria`	string	Criterion wording
`criteria.<code>.thoughts`	string	Justification of the score

Response example

The response to the request from the examples above. Base call fields always arrive, and the AI blocks contain exactly what is listed in select: inside overview — only the topic, the agreements and the action items, inside insights — only the participant analysis, inside evaluation — only the overall score. Employee mentions arrive as plain text, because mentionFormat: "none" was requested.

JSON

{
  "success": true,
  "data": {
    "items": [
      {
        "callId": 12345,
        "callType": 1,
        "initiatorId": 7,
        "startDate": "2026-01-15T10:00:00+00:00",
        "endDate": "2026-01-15T10:42:00+00:00",
        "durationSeconds": 2520,
        "participants": [
          { "userId": 7, "name": "John Smith", "avatar": "https://example.bitrix24.com/upload/avatar.png", "workPosition": "Project manager", "talkedSeconds": 600 },
          { "userId": 42, "name": "Mary Jones", "avatar": "https://example.bitrix24.com/upload/avatar-2.png", "workPosition": "Developer", "talkedSeconds": 1200 }
        ],
        "outcomes": ["transcription", "overview", "summary", "insights", "evaluation"],
        "overview": {
          "topic": "Sprint planning",
          "agreements": [
            { "agreement": "Only the catalog import goes into the sprint", "quote": "I suggest we keep only the catalog import, we will not finish the rest." }
          ],
          "actionItems": [
            { "actionItem": "Prepare the prototype by Friday", "quote": "Then we show the prototype on Friday." }
          ]
        },
        "insights": {
          "speakerAnalysis": [
            { "userId": 42, "detailedInsight": "The participant proposed cutting the scope and backed it with a deadline estimate.", "efficiencyValue": 82, "evaluationCriteria": "Constructiveness of proposals", "talkPercentage": 66, "duration": 1200, "durationFormat": "20:00" },
            { "userId": 7, "detailedInsight": "The participant ran the meeting and recorded the agreements.", "efficiencyValue": 74, "evaluationCriteria": "Discussion management", "talkPercentage": 34, "duration": 600, "durationFormat": "10:00" }
          ]
        },
        "evaluation": {
          "efficiencyValue": 75
        }
      },
      {
        "callId": 12318,
        "callType": 2,
        "initiatorId": 42,
        "startDate": "2026-01-12T09:00:00+00:00",
        "endDate": "2026-01-12T09:25:00+00:00",
        "durationSeconds": 1500,
        "participants": [
          { "userId": 42, "name": "Mary Jones", "avatar": "https://example.bitrix24.com/upload/avatar-2.png", "workPosition": "Developer", "talkedSeconds": 900 }
        ],
        "outcomes": ["transcription", "overview"],
        "overview": {
          "topic": "Review of the week's support requests",
          "agreements": [],
          "actionItems": [
            { "actionItem": "Collect frequent questions into the knowledge base", "quote": "Let's collect the recurring questions into one article." }
          ]
        },
        "insights": null,
        "evaluation": null
      }
    ],
    "hasMore": true,
    "afterCursor": { "startDate": "2026-01-12T09:00:00.000000+00:00", "id": 12318 }
  }
}

The second call shows a common case: not all of its AI blocks are ready. outcomes lists what has been generated, and blocks that were requested but are absent arrive as null — check them before accessing their fields.

See the Follow-up for a single call page for a call with all of its blocks filled in.

Error response example

400 — the request body has no filter object:

JSON

{
  "success": false,
  "error": {
    "code": "MISSING_PARAMS",
    "message": "Required: filter.startDate.from/to (ISO 8601)"
  }
}

Errors

HTTP	Code	Description
400	`MISSING_PARAMS`	The request body has no `filter` object
400	`INVALID_PARAMS`	The value of a request field is not in the list of allowed values — `mentionFormat`, for example. If Bitrix24 rejected the value of a specific field, the response additionally contains an `error.validation` array with the field name
401	`MISSING_API_KEY`	The `X-Api-Key` header was not passed
401	`TOKEN_MISSING`	The key has no portal tokens. An OAuth application key requires the `Authorization: Bearer` header
403	`SCOPE_DENIED`	The key has no `call` scope
403	`BITRIX_ACCESS_DENIED`	Bitrix24 denied access: the key's webhook has no `call` permission on the portal side. Reconnect the key with the required permission
422	`METHOD_NOT_YET_AVAILABLE`	Update `call 26.600.0` has not been released on the portal yet. The response contains an `error.release` field with the target version
422	`BITRIX_ERROR`	The request was rejected: incorrect date range, disallowed field in `select`, incorrect `pagination` or `order`, no access to the data. The reason is in `error.message`
429	`RATE_LIMITED`	Request rate exceeded on the Bitrix24 side
429	`QUEUE_OVERFLOW`, `QUEUE_TIMEOUT`	The portal request queue is full or the request did not get through the queue in time. The `Retry-After` header suggests the delay before a retry
503	`BITRIX_TIMEOUT`	Bitrix24 accepted the request but did not respond within 15 seconds. For reads a retry is safe
502	`BITRIX_UNAVAILABLE`	Bitrix24 is unavailable

The full list of common API errors — Errors.

Known specifics

Calls without AI processing do not appear in the list. The output contains only the calls for which a Follow-up has already been generated. An empty items array means there are no such calls for the period.

An incomplete period is rejected with code 422, not 400. The code 400 MISSING_PARAMS is returned only when there is no filter object in the request body at all. A missing to, or a period where from is later than to, returns 422 BITRIX_ERROR with a description of the reason.

Participant analysis is not available on every Bitrix24 account. When it is unavailable, the insights block comes back with speakerEvaluationAvailable: false and an empty participant analysis. This is not a request error — the response structure does not change.

The output is limited by the employee's permissions. The key returns Follow-ups of the calls in which the key owner took part, or whose linked chat the key owner belongs to. The Bitrix24 account administrator sees all calls of the account. That is why the same period gives different lists for different employees.

#Call follow-up

#Follow-up list

#Request fields (body)

#Select field

#Metadata

#AI blocks

#Navigation

#Mention format

#Examples

#curl — personal key

#curl — OAuth application

#JavaScript — personal key

#JavaScript — OAuth application

#Response fields

#Participants — `participants[]`

#Call recordings — `tracks[]`

#Transcription — `transcription`

#Meeting overview — `overview`

#Summary — `summary`

#Insights — `insights`

#Efficiency score — `evaluation`

#Response example

#Error response example

#Errors

#Known specifics

#See also