> ## Documentation Index
> Fetch the complete documentation index at: https://docs.vook.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Ask a question about a transcript and get the answer

> You send a free-text prompt that runs against the transcription and get the answer back in the same response. The exchange is saved as a chat so you can list and retrieve it later. Next: GET the chat by its id to read the answer again.



## OpenAPI

````yaml /api-reference/openapi.json post /api/v1/transcriptions/{id}/chats
openapi: 3.0.0
info:
  title: Vook API
  description: >-
    Public transcription API. Authenticate every request with your API key:
    `Authorization: Bearer vk_live_…`.


    ## Transcribe a file (3 calls)


    1. **`POST /api/v1/uploads/init`** → returns an `upload_token`.

    2. **`POST {ingress}/upload`**: upload the whole file to the ingress host
    (`https://ingress.vook.ai`, **a different host from this API**) as
    `multipart/form-data` with a single `chunk` field, authenticated with
    `Authorization: Bearer {upload_token}`. The file is sent as a single chunk;
    no chunk index/count is needed. This step is on the ingress service and is
    not listed below.

    3. **`POST /api/v1/transcription-jobs`**: pass the `upload_token` +
    `file_name` + `language` + `diarize`. Returns a job `id`.


    Then **poll `GET /api/v1/transcription-jobs/{id}`** until `status` is
    `completed` (or `partial`/`failed`). Read each id in `transcription_ids`
    through **`GET /api/v1/transcriptions/{id}`** and its `/transcript` /
    `/export` sub-resources.
  version: v1
  contact: {}
servers:
  - url: https://www.api.vook.ai
    description: Production
security: []
tags: []
paths:
  /api/v1/transcriptions/{id}/chats:
    post:
      tags:
        - Chats (API v1)
      summary: Ask a question about a transcript and get the answer
      description: >-
        You send a free-text prompt that runs against the transcription and get
        the answer back in the same response. The exchange is saved as a chat so
        you can list and retrieve it later. Next: GET the chat by its id to read
        the answer again.
      operationId: ApiV1LlmChatController_create
      parameters:
        - name: id
          required: true
          in: path
          description: Transcription ID
          example: 123e4567-e89b-12d3-a456-426614174000
          schema:
            type: string
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ApiCreateChatDto'
      responses:
        '201':
          description: The created chat, including the answer
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ApiChatDto'
      security:
        - bearer: []
components:
  schemas:
    ApiCreateChatDto:
      type: object
      properties:
        title:
          type: string
          minLength: 1
          maxLength: 255
          description: Short label for this exchange, shown when you list chats
          example: Action items
        prompt:
          type: string
          minLength: 1
          description: Question or instruction to run against the transcript
          example: List the action items decided in this meeting.
      required:
        - title
        - prompt
    ApiChatDto:
      type: object
      properties:
        id:
          type: string
          description: Chat ID
          example: 123e4567-e89b-12d3-a456-426614174000
        transcription_id:
          type: string
          description: ID of the transcription this chat runs against
          example: 123e4567-e89b-12d3-a456-426614174000
        title:
          type: string
          description: Label you gave the chat
          example: Action items
        prompt:
          type: string
          nullable: true
          description: The prompt you sent
          example: List the action items decided in this meeting.
        answer:
          type: string
          nullable: true
          description: The answer produced over the transcript
          example: 1. Ship the API docs. 2. Review the pricing page.
        created_at:
          format: date-time
          type: string
          description: When the chat was created
          example: '2024-01-15T10:30:00.000Z'
      required:
        - id
        - transcription_id
        - title
        - created_at

````