International Business Machines Text Generation API

openapi: 3.2.0
info:
  title: International Business Machines IBM watsonx.ai Text Generation API
  description: The IBM watsonx.ai API enables developers to run text inference, prompt tuning, and other operations on Large Language Models (LLMs). It provides access to multiple open source and IBM foundation models through a unified REST interface for text generation, embeddings, and model management.
  version: 1.0.0
  contact:
    name: IBM Cloud
    url: https://cloud.ibm.com/apidocs/watsonx-ai
  license:
    name: IBM Cloud Terms
    url: https://www.ibm.com/terms
servers:
- url: https://us-south.ml.cloud.ibm.com
  description: US South (Dallas)
- url: https://eu-de.ml.cloud.ibm.com
  description: EU Central (Frankfurt)
- url: https://eu-gb.ml.cloud.ibm.com
  description: EU (London)
- url: https://jp-tok.ml.cloud.ibm.com
  description: AP (Tokyo)
security:
- bearerAuth: []
tags:
- name: Text Generation
  description: Generate text using foundation models.
paths:
  /ml/v1/text/generation:
    post:
      operationId: textGeneration
      summary: International Business Machines Generate text
      description: Infer the next tokens for a given deployed model with a set of parameters.
      tags:
      - Text Generation
      parameters:
      - name: version
        in: query
        required: true
        schema:
          type: string
        description: The API version date.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TextGenerationRequest'
      responses:
        '200':
          description: Successful response
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TextGenerationResponse'
        '400':
          description: Bad request
        '401':
          description: Unauthorized
  /ml/v1/text/generation_stream:
    post:
      operationId: textGenerationStream
      summary: International Business Machines Generate text (streaming)
      description: Infer the next tokens for a given deployed model with a set of parameters, returning results as a stream.
      tags:
      - Text Generation
      parameters:
      - name: version
        in: query
        required: true
        schema:
          type: string
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TextGenerationRequest'
      responses:
        '200':
          description: Successful streaming response
          content:
            text/event-stream:
              schema:
                type: string
  /ml/v1/text/tokenization:
    post:
      operationId: textTokenization
      summary: International Business Machines Tokenize text
      description: Return the tokens for the provided text with the given model.
      tags:
      - Text Generation
      parameters:
      - name: version
        in: query
        required: true
        schema:
          type: string
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/TokenizationRequest'
      responses:
        '200':
          description: Successful response
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/TokenizationResponse'
components:
  schemas:
    TextGenerationResponse:
      type: object
      properties:
        model_id:
          type: string
        created_at:
          type: string
        results:
          type: array
          items:
            type: object
            properties:
              generated_text:
                type: string
              generated_token_count:
                type: integer
              input_token_count:
                type: integer
              stop_reason:
                type: string
    TextGenerationRequest:
      type: object
      required:
      - model_id
      - input
      - project_id
      properties:
        model_id:
          type: string
          description: The ID of the model to use.
        input:
          type: string
          description: The prompt text.
        project_id:
          type: string
          description: The project ID.
        parameters:
          type: object
          properties:
            max_new_tokens:
              type: integer
            temperature:
              type: number
            top_p:
              type: number
            top_k:
              type: integer
            repetition_penalty:
              type: number
            stop_sequences:
              type: array
              items:
                type: string
    TokenizationRequest:
      type: object
      required:
      - model_id
      - input
      - project_id
      properties:
        model_id:
          type: string
        input:
          type: string
        project_id:
          type: string
    TokenizationResponse:
      type: object
      properties:
        model_id:
          type: string
        result:
          type: object
          properties:
            token_count:
              type: integer
            tokens:
              type: array
              items:
                type: string
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer
      bearerFormat: IAM Token