Pryon Retrieval API

The Retrieval API from Pryon — 4 operation(s) for retrieval.

Work with this as data

Every API here is available over the APIs.io API and to AI agents over MCP.

MCP server

One button, every client — Claude, Cursor, VS Code and the rest.

https://apis.io/mcp

Tools for apis

7 MCP tools reach this
  • find_apisBrowse and filter every API in the catalog.
  • get_api_artifactsOne API's artifacts, grouped by type.
  • get_openapiThe primary OpenAPI for this API.
  • find_similar_apisAPIs that look like this one.
  • apis_io_searchSTART HERE — APIs, providers and tags for one query, each with its total.
  • resolveTurn a domain, URL or GitHub org into the provider it belongs to.
  • find_cohortsEvery scored population of providers in the catalog.
All 92 tools

Call it yourself

curl for this page
This API
curl "https://apis.io/api/v1/apis/pryon-retrieval-api"
All apis
curl "https://apis.io/api/v1/apis?limit=25"

Discovery needs no key. Ratings and market analysis are Pro.

Get an API key

Free tier, no email required.

A second provider on the same verified email joins the account you already have.

OpenAPI Specification

pryon-retrieval-api-openapi.yml Raw ↑
openapi: 3.2.0
info:
  title: Retrieval API
  version: '1.0'
  contact:
    name: Pryon Product Support
    url: https://pryon.zendesk.com
    email: support@pryon.zendesk.com
  description: ''
  x-logo:
    altText: Pryon APIs
servers:
- url: https://api.pryon.net
  description: Pryon Cloud
tags:
- name: Retrieval
paths:
  /api/conversation/v1alpha1/retrievals:
    post:
      operationId: Retrieval_CreateRetrieval
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/filename_v1alpha1CreateRetrievalRequest'
        required: true
      responses:
        '200':
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/filename_v1alpha1GetRetrievalResponse'
          description: A successful response.
        default:
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/filename_googlerpcStatus'
          description: An unexpected error response.
      summary: Create Retrieval
      tags:
      - Retrieval
      x-codegen-request-body-name: body
      description: Initiate or continue a conversation with the retrieval service.
      parameters:
      - name: Authorization
        in: header
        required: true
        schema:
          type: string
        example: 'Authorization: Bearer {base64-encoded-token}'
  /api/conversation/v1alpha1/retrievals/wss:
    post:
      description: Initiate or continue a conversation.
      operationId: Retrieval_CreateWsRetrieval
      requestBody:
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/filename_v1alpha1CreateWsRetrievalRequest'
        required: true
      responses:
        '200':
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/filename_v1alpha1WsRetrieval'
          description: A successful response.
        default:
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/filename_googlerpcStatus'
          description: An unexpected error response.
      summary: Create Ws Retrieval
      tags:
      - Retrieval
      x-codegen-request-body-name: body
      parameters:
      - name: Authorization
        in: header
        required: true
        schema:
          type: string
        example: 'Authorization: Bearer {base64-encoded-token}'
  /api/conversation/v1alpha1/retrievals/{retrieval_id}:
    get:
      operationId: Retrieval_GetRetrieval
      parameters:
      - description: Unique identifier associated with the response object sources.
        in: path
        name: retrieval_id
        required: true
        schema:
          type: string
      - description: Unique identifier for the output.
        in: query
        name: output_id
        schema:
          type: string
      - name: Authorization
        in: header
        required: true
        schema:
          type: string
        example: 'Authorization: Bearer {base64-encoded-token}'
      responses:
        '200':
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/filename_v1alpha1GetRetrievalResponse'
          description: A successful response.
        default:
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/filename_googlerpcStatus'
          description: An unexpected error response.
      summary: Get Retrieval
      tags:
      - Retrieval
      description: Retrieve the query and output data for existing retrieval data.
  /api/conversation/v1alpha1/retrievals/{retrieval_id}/image:
    get:
      description: Get an image representation of the content with bounding boxes.
      operationId: Retrieval_GetRetrievalContentImage
      parameters:
      - description: Unique identifier for the response object for the bounding boxes and content ID.
        in: path
        name: retrieval_id
        required: true
        schema:
          type: string
      - description: Unique identifier for the retrieval output; defaults to the first instance.
        in: query
        name: output_id
        schema:
          type: string
      - description: Page number in the content; defaults to the first content page associated with the retrieval output.
        in: query
        name: page
        schema:
          format: int32
          type: integer
      - description: Compress image data; valid values are 0 for compressed (default) and 1 for uncompressed.
        in: query
        name: uncompressed
        schema:
          type: boolean
      - description: Bounding box overlay color specified as a hex color code without the leading hashtag; supports 3, 4, 6, and 8 character codes.
        in: query
        name: overlay_color
        schema:
          type: string
      - name: Authorization
        in: header
        required: true
        schema:
          type: string
        example: 'Authorization: Bearer {base64-encoded-token}'
      responses:
        '200':
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/filename_v1alpha1GetRetrievalContentImageResponse'
          description: A successful response.
        default:
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/filename_googlerpcStatus'
          description: An unexpected error response.
      summary: Get Retrieval Content Image
      tags:
      - Retrieval
components:
  schemas:
    filename_v1alpha1BoundingBox:
      properties:
        x1:
          format: float
          title: Required. top-left corner X-axis
          type: number
        y1:
          format: float
          title: Required. top-left corner Y-axis
          type: number
        x2:
          format: float
          title: Bottom-right corner X-axis.
          type: number
        y2:
          format: float
          title: Bottom-right corner Y-axis.
          type: number
        page_number:
          format: int32
          title: Page number where the bounding box was found.
          type: integer
        page_width:
          format: float
          title: The width of the page the bounding box was found on
          type: number
        page_height:
          format: float
          title: The height of the page the bounding box was found on
          type: number
      title: A rectangle bounding box typically for indicating where text matched on a page
      type: object
    filename_v1alpha1RetrievalFilter:
      properties:
        collection_id:
          description: Parent collection for the knowledge domain; autogenerated and not editable.
          type: string
        knowledge_domain_id:
          description: Unique identifier of the knowledge domain that supplied the exchange response.
          type: string
        content_group_ids:
          description: Unique identifier for the content group.
          items:
            type: string
          type: array
        tags:
          description: Tags used to return data.
          items:
            $ref: '#/components/schemas/filename_v1alpha1TagData'
          type: array
      title: Filters for limiting what content is searched
      type: object
    filename_v1alpha1ContentMetadata:
      properties:
        content_id:
          description: Unique identifier for the content within the matching content groups.
          type: string
        content_display_name:
          description: Display name of the source file that includes the best sentence answer.
          type: string
        content_source_uri:
          description: Unique identifier for the source content that contains the answer.
          type: string
        content_origin_source_url:
          description: URL for the original source location of a document.
          type: string
        content_create_time:
          format: date-time
          title: Time of the initial content update.
          type: string
        content_update_time:
          format: date-time
          title: Time of the last update; 0 indicates updates have never been made.
          type: string
        image_uri:
          title: First page of the document where the answer was found.
          type: string
        hyperlinks:
          items:
            type: string
          title: Relevant links to the web-based document.
          type: array
        author:
          title: Name of the author who created the document.
          type: string
        author_id:
          title: Unique identifier for the author who created the document.
          type: string
        title:
          description: Name or title of the document; maximum size is 255 characters.
          type: string
        description:
          description: Summary of the document; maximum size is 1000 characters.
          type: string
        mime_type:
          title: Document media type.
          type: string
        additional_metadata:
          additionalProperties:
            properties: {}
            type: object
          title: Metadata captured when ingesting the document.
          type: object
      title: Additional metadata from the source content
      type: object
    filename_RetrievalOutputAnswerType:
      default: UNSPECIFIED
      description: "- UNSPECIFIED: None\n - TEXT_EXCERPT: Default\n - FAQ: Answer comes from FAQ.\n - VA_PROMOTED: Promoted answer from existing content.\n - VA_CURATED: Verified answer created as standalone VA content.\n - BELOW_CONFIDENCE_THRESHOLD: Results scored below a minimum confidence threshold.\n - NO_ANSWER: No answer matched, or no permissions to see the matches; alternate answer supplied.\n - INTELLIGIBILITY_OUT_OF_SCOPE: Query is classified as gibberish and is not intelligible.\n - LANGUAGE_OUT_OF_SCOPE: Query is not in a supported language.\n - CONVERSATIONAL_OUT_OF_SCOPE: Query is classified as conversational."
      title: General source of the answer
      type: string
    filename_protobufAny:
      properties:
        type_url:
          type: string
        value:
          format: byte
          pattern: ^(?:[A-Za-z0-9+/]{4})*(?:[A-Za-z0-9+/]{2}==|[A-Za-z0-9+/]{3}=)?$
          type: string
      type: object
    filename_v1alpha1WsRetrieval:
      description: The only top-level message returned to the client by the CreateWsRetrieval (WebSocket) method.
      properties:
        metadata:
          $ref: '#/components/schemas/filename_v1alpha1ResourceMetadata'
        user_id:
          title: Unique identifier of the user who made the retrieval.
          type: string
        knowledge_domain_id:
          title: Unique identifier of the knowledge domain that supplied the exchange response.
          type: string
        collection_id:
          title: Unique identifier of the knowledge collection where the response was located.
          type: string
        state:
          $ref: '#/components/schemas/filename_WsRetrievalState'
        retrieval:
          $ref: '#/components/schemas/filename_v1alpha1RetrievalData'
        error:
          $ref: '#/components/schemas/filename_googlerpcStatus'
      type: object
    filename_v1alpha1ResourceMetadata:
      properties:
        uuid:
          description: Unique identifier for the interaction returned in the resource response; the response ID.
          type: string
        create_time:
          description: Time of the initial request.
          format: date-time
          type: string
        update_time:
          description: Time of the last update; 0 indicates updates have never been made.
          format: date-time
          type: string
        response_time_millis:
          format: int64
          description: Response time (in milliseconds).
          type: integer
      type: object
    filename_v1alpha1GetRetrievalContentImageResponse:
      title: GetRetrievalContentImage returns the image bytes directly in the REST response, and not through gRPC
      type: object
    filename_v1alpha1CreateRetrievalRequest:
      description: Message sent by the client for the create retrieval method.
      properties:
        conversation_id:
          description: Unique identifier for the conversation returned with the create retrieval response; leaving blank starts a new conversation on the request.
          type: string
        language_id:
          description: Language tags; default is en-US.
          type: string
        query:
          description: Input query used for retrieval.
          type: string
        audio:
          $ref: '#/components/schemas/filename_v1alpha1RetrievalAudioInput'
        options:
          $ref: '#/components/schemas/filename_v1alpha1RetrievalOption'
        filters:
          $ref: '#/components/schemas/filename_v1alpha1RetrievalFilter'
      type: object
    filename_v1alpha1ContentGroupSparse:
      properties:
        content_group_id:
          title: Unique identifier for the content group.
          type: string
        display_name:
          title: Name of the collection.
          type: string
      title: Abbreviated form of ContentGroup for returning only the essential information.
      type: object
    filename_CreateWsRetrievalRequestAction:
      default: UNSPECIFIED
      description: "- UNSPECIFIED: This indicates that an action was not specified, as it is the protobuf default.  This is never a valid value, and if received, then an ERROR response message will be returned, and the connection will be closed by the server.\n - START: START: signals the intent to begin a Retrieval. If an existing conversation is being continued, then conversation_id should be set to the value used in previous Retrievals.\nIf text input mode is being used, this text-based message must also include the complete input query.\nIf audio input mode is being used, this text-based message must include valid input.raw_audio.audio_encoding and input.raw_audio.sample_rate_hertz fields. It cannot include input.raw_audio.content, as input audio should be sent in binary messages that follow. If invalid values are requested for audio_encoding and/or sample_rate_hertz, then an ERROR response message will be sent, and the connection will be closed by the server.\nIn both text and audio input modes, a successful start request will be acknowledged by the server with a PROCESSING text message. Frames of audio should be sent by the client as binary messages following receipt of this PROCESSING text message.  After all frames of audio have been sent, a STOP message should be sent by the client.\n - STOP: STOP: This message should be sent following the START message. When audio input mode is being used, STOP should be sent by the client after all audio frames have been sent to the server."
      title: Action
      type: string
    filename_v1alpha1RetrievalAudioOutputAudioEncoding:
      default: UNSPECIFIED
      description: "- UNSPECIFIED: Not specified.\n - LINEAR16: LINEAR16: Uncompressed 16-bit signed little-endian samples (Linear PCM).\n - MP3: MP3: MP3 audio encoding. The sample rate is encoded in the payload."
      title: Audio encoding of the content in the message
      type: string
    filename_v1alpha1TagData:
      properties:
        source:
          $ref: '#/components/schemas/filename_v1alpha1TagSourceType'
        type:
          title: A grouping of related types of tags
          type: string
        value:
          title: The most specific part of the tag
          type: string
      type: object
    filename_v1alpha1RetrievalOutput:
      properties:
        id:
          description: Unique identifier for the output resource.
          type: string
        answer_snippet:
          description: Best matching snippet displayed to the end-user
          type: string
        final_query:
          description: Specific matched query after synonym expansion and prior knowledge corrections.
          type: string
        context_chunk:
          description: Full surrounding text content for the matched text.
          type: string
        context_chunk_start_page:
          description: Page number where the answer in context starts in a document.
          format: int32
          type: integer
        context_chunk_end_page:
          description: Page number where the answer in context ends in a document.
          format: int32
          type: integer
        context_chunk_bbox:
          items:
            $ref: '#/components/schemas/filename_v1alpha1BoundingBox'
          description: Bounding boxes of answer in context; generally one response but can be multiple if the context chunk spans pages.
          type: array
        score:
          description: Approximate strength of the returned answer in context.
          format: float
          type: number
        answer_snippet_score:
          format: float
          description: Approximate gauge of how well the the answer snippet matched the query.
          type: number
        answer_snippet_start_index:
          description: Character index of the first occurrence of the answer text within the surrounding answer in context.
          format: int32
          type: integer
        answer_snippet_end_index:
          description: Character index of the last occurrence of the answer within the surrounding answer in context.
          format: int32
          type: integer
        answer_snippet_bbox:
          items:
            $ref: '#/components/schemas/filename_v1alpha1BoundingBox'
          title: Bounding boxes for each line of the text match.
          type: array
        audio:
          $ref: '#/components/schemas/filename_v1alpha1RetrievalAudioOutput'
        confidence_label:
          description: Answer confidence levels; used to categorize outputs.
          type: string
        tags:
          description: Tags associated with the output.
          items:
            $ref: '#/components/schemas/filename_v1alpha1TagData'
          type: array
        answer_type:
          $ref: '#/components/schemas/filename_RetrievalOutputAnswerType'
        verified_answer_id:
          description: Unique identifier of the verified answer; if the answer isn't verified this is unset.
          type: string
        video_time_offset_seconds:
          description: Length of time from the beginning of a in video (in seconds) until the short answer is found.
          format: int32
          type: integer
        content_group:
          $ref: '#/components/schemas/filename_v1alpha1ContentGroupSparse'
        content_metadata:
          $ref: '#/components/schemas/filename_v1alpha1ContentMetadata'
        context:
          $ref: '#/components/schemas/filename_v1alpha1RetrievalContext'
        answer_html:
          title: A marked up HTML version of the text_wordy/answer_in_context text
          type: string
      title: Each Retrieval output represents text matched in source content against the input
      type: object
    filename_WsRetrievalState:
      default: UNSPECIFIED
      description: "- UNSPECIFIED: Not specified.\n - ERROR: Unsupported request received or problem with the API server.\n - PROCESSING: START request for audio input successful.\n - QUERY_RESULT: API returned the Retrieval response to the client.\n - PROCESSING_COMPLETED: Full response sent for retrieval request."
      title: State of the Retrieval
      type: string
    filename_v1alpha1NormalizedRetrievalInput:
      properties:
        normalized_input_id:
          description: Unique identifier for the normalized input resource.
          type: string
        query:
          description: Original input query.
          type: string
        understood_tags:
          description: Tags detected from the query and used for filtering.
          items:
            $ref: '#/components/schemas/filename_v1alpha1TagData'
          type: array
      title: Post-processed version of the input
      type: object
    filename_v1alpha1RetrievalContext:
      description: Additional context for the output.
      properties:
        augmentation:
          description: Predefined additional information provided with the source of this output.
          properties: {}
          type: object
      type: object
    filename_v1alpha1RetrievalAudioInputAudioEncoding:
      description: "- UNSPECIFIED: Encoding not specified; default and returns an ERROR response.\n - LINEAR16: Uncompressed 16-bit signed little-endian samples (Linear PCM)."
      title: Audio encoding of the content in the message; must be one-channel (mono).
      type: string
    filename_v1alpha1RetrievalData:
      properties:
        user_id:
          title: The identifier of the user that originally made the Retrieval
          type: string
        conversation_id:
          description: Unique identifier for conversations for subsequent calls; saved in the client and returned on the next request.
          type: string
        retrieval_id:
          description: Unique ID of the response item.
          type: string
        collection_id:
          description: Parent collection for the knowledge domain; autogenerated and not editable.
          type: string
        knowledge_domain_id:
          description: Unique identifier of the knowledge domain that supplied the exchange response.
          type: string
        normalized_input:
          $ref: '#/components/schemas/filename_v1alpha1NormalizedRetrievalInput'
        suggested_query:
          description: Spelling corrections, if any, or other suggestions found for the input query.
          items:
            type: string
          type: array
        outputs:
          description: List of outputs answering the question ported from retrieval result.
          items:
            $ref: '#/components/schemas/filename_v1alpha1RetrievalOutput'
          type: array
      title: A Retrieval is the primary object for getting content matched against the question asked in CreateRetrieval
      type: object
    filename_v1alpha1RetrievalAudioInput:
      properties:
        audio_encoding:
          $ref: '#/components/schemas/filename_v1alpha1RetrievalAudioInputAudioEncoding'
        sample_rate_hertz:
          description: Sample rate in hertz of the audio data; only valid value is 16000, which must be explicitly stated by the client.
          format: int32
          type: integer
        content:
          description: Bytes of encoded audio data.
          format: byte
          pattern: ^(?:[A-Za-z0-9+/]{4})*(?:[A-Za-z0-9+/]{2}==|[A-Za-z0-9+/]{3}=)?$
          type: string
      title: Audio input for speech to text
      type: object
    filename_v1alpha1GetRetrievalResponse:
      properties:
        metadata:
          $ref: '#/components/schemas/filename_v1alpha1ResourceMetadata'
        data:
          $ref: '#/components/schemas/filename_v1alpha1RetrievalData'
      type: object
    filename_googlerpcStatus:
      properties:
        code:
          format: int32
          type: integer
        message:
          type: string
        details:
          items:
            $ref: '#/components/schemas/filename_protobufAny'
          type: array
      type: object
    filename_v1alpha1RetrievalAudioOutput:
      properties:
        audio_output_id:
          description: Unique identifier for the audio output resource.
          type: string
        audio_encoding:
          $ref: '#/components/schemas/filename_v1alpha1RetrievalAudioOutputAudioEncoding'
        sample_rate_hertz:
          description: Sample rate in hertz of the audio data; only valid value is 16000, which must be explicitly stated by the client.
          format: int32
          type: integer
        content:
          description: Bytes of encoded audio data.
          format: byte
          pattern: ^(?:[A-Za-z0-9+/]{4})*(?:[A-Za-z0-9+/]{2}==|[A-Za-z0-9+/]{3}=)?$
          type: string
      title: Audio output when answer is converted back to speech
      type: object
    filename_v1alpha1TagSourceType:
      default: TAG_SOURCE_TYPE_UNSPECIFIED
      description: " - TAG_SOURCE_TYPE_UNSPECIFIED: Default. No tags can be created with this source.\n - SYSTEM_DEFINED: Tags created by the Pryon Cloud Services system.\n - USER_DEFINED: Tags defined by users if the system and created via user input.\n - QUERY_ATTRIBUTE: Tags associated to a content which can be used to improve Exchange response ranking quality based on the context specified in ExchangeInputOption."
      type: string
    filename_v1alpha1CreateWsRetrievalRequest:
      description: The top-level message sent by the client for the CreateWsRetrieval (WebSocket) method.
      properties:
        ws_retrieval_id:
          description: Unique identifier for the retrieval item.
          type: string
        action:
          $ref: '#/components/schemas/filename_CreateWsRetrievalRequestAction'
        Retrieval_request:
          $ref: '#/components/schemas/filename_v1alpha1CreateRetrievalRequest'
      type: object
    filename_v1alpha1RetrievalOption:
      properties:
        audio_output_enabled:
          description: Returns audio output when set to True; if no audio output is generated, responses include only text output; default is false.
          type: boolean
        max_outputs:
          description: Number of outputs to be returned; minimum is 3 (default), maximum is 10.
          format: int32
          type: integer
        context:
          properties: {}
          title: Additional context for the exchange input; improves response ranking quality.
          type: object
      title: Additional options for customizing the result of a Retrieval
      type: object
  securitySchemes:
    oauth2_token:
      description: The Pryon APIs are authenicated with an OAuth2 Bearer token.
      type: apiKey
      name: 'Authoriation: Bearer'
      in: header
x-tagGroups:
- name: Retrieval APIs
  tags:
  - Content Service
  - Exchange
  - Glossary