Vectara Corpora API

openapi: 3.1.0
info:
  title: Vectara REST Agents Corpora API
  description: 'Vectara REST API v2 is the unified interface for the Vectara Retrieval Augmented

    Generation (RAG) platform. It exposes endpoints for managing corpora, uploading

    and indexing documents, running semantic and hybrid queries with grounded

    generation, managing agents, tools, pipelines, generation presets, and reading

    metadata.


    This OpenAPI 3.1 description is hand-curated from the public Vectara REST API

    documentation at https://docs.vectara.com/docs/rest-api/ and the published

    OpenAPI document at https://docs.vectara.com/vectara-oas-v2.yaml. It covers a

    representative subset of endpoints across corpora, documents, query, agents,

    tools, pipelines, and OAuth.

    '
  version: '2.0'
  contact:
    name: Vectara Support
    url: https://docs.vectara.com/docs/support
  license:
    name: Proprietary
servers:
- url: https://api.vectara.io
  description: Vectara REST API production base URL
- url: https://auth.vectara.io
  description: Vectara OAuth 2.0 token endpoint
security:
- ApiKeyAuth: []
- OAuth2: []
tags:
- name: Corpora
  description: Create, list, update, and delete corpora that hold indexed documents.
paths:
  /v2/corpora:
    get:
      tags:
      - Corpora
      summary: List Corpora
      operationId: listCorpora
      parameters:
      - in: query
        name: limit
        schema:
          type: integer
          default: 10
      - in: query
        name: page_key
        schema:
          type: string
      - in: query
        name: filter
        schema:
          type: string
      responses:
        '200':
          description: A list of corpora.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ListCorporaResponse'
    post:
      tags:
      - Corpora
      summary: Create A Corpus
      operationId: createCorpus
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/CreateCorpusRequest'
      responses:
        '201':
          description: Corpus created.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Corpus'
  /v2/corpora/{corpus_key}:
    parameters:
    - $ref: '#/components/parameters/CorpusKey'
    get:
      tags:
      - Corpora
      summary: Retrieve Corpus Metadata
      operationId: getCorpus
      responses:
        '200':
          description: Corpus metadata.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Corpus'
    patch:
      tags:
      - Corpora
      summary: Update A Corpus
      operationId: updateCorpus
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/UpdateCorpusRequest'
      responses:
        '200':
          description: Corpus updated.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Corpus'
    delete:
      tags:
      - Corpora
      summary: Delete A Corpus
      operationId: deleteCorpus
      responses:
        '204':
          description: Corpus deleted.
  /v2/corpora/{corpus_key}/reset:
    parameters:
    - $ref: '#/components/parameters/CorpusKey'
    post:
      tags:
      - Corpora
      summary: Reset A Corpus
      description: Remove all documents from the corpus while keeping the corpus itself.
      operationId: resetCorpus
      responses:
        '204':
          description: Corpus reset.
  /v2/corpora/{corpus_key}/compute_size:
    parameters:
    - $ref: '#/components/parameters/CorpusKey'
    post:
      tags:
      - Corpora
      summary: Compute Corpus Size
      operationId: computeCorpusSize
      responses:
        '200':
          description: Corpus size information.
          content:
            application/json:
              schema:
                type: object
                properties:
                  size_bytes:
                    type: integer
                  document_count:
                    type: integer
components:
  schemas:
    PageMetadata:
      type: object
      properties:
        page_key:
          type: string
        next_page_key:
          type: string
    CreateCorpusRequest:
      allOf:
      - $ref: '#/components/schemas/Corpus'
      - type: object
        required:
        - key
        - name
    UpdateCorpusRequest:
      type: object
      properties:
        name:
          type: string
        description:
          type: string
        enabled:
          type: boolean
    FilterAttribute:
      type: object
      properties:
        name:
          type: string
        level:
          type: string
          enum:
          - document
          - document_part
        indexed:
          type: boolean
        type:
          type: string
          enum:
          - text
          - integer
          - real
          - boolean
    ListCorporaResponse:
      type: object
      properties:
        corpora:
          type: array
          items:
            $ref: '#/components/schemas/Corpus'
        metadata:
          $ref: '#/components/schemas/PageMetadata'
    Corpus:
      type: object
      properties:
        key:
          type: string
        name:
          type: string
        description:
          type: string
        queries_are_answers:
          type: boolean
        documents_are_questions:
          type: boolean
        encoder_id:
          type: string
        filter_attributes:
          type: array
          items:
            $ref: '#/components/schemas/FilterAttribute'
        custom_dimensions:
          type: array
          items:
            type: object
            properties:
              name:
                type: string
              description:
                type: string
        enabled:
          type: boolean
        created_at:
          type: string
          format: date-time
  parameters:
    CorpusKey:
      in: path
      name: corpus_key
      required: true
      schema:
        type: string
      description: The unique key identifying a corpus.
  securitySchemes:
    ApiKeyAuth:
      type: apiKey
      in: header
      name: x-api-key
      description: Vectara API key passed in the `x-api-key` header.
    OAuth2:
      type: oauth2
      description: OAuth 2.0 client credentials flow.
      flows:
        clientCredentials:
          tokenUrl: https://auth.vectara.io/oauth2/token
          scopes: {}