openapi: 3.1.0
info:
  title: Qubrid AI Compute Chat Completions API
  description: The Qubrid AI Compute API provides programmatic access to GPU cloud infrastructure including NVIDIA H100, H200, and B200 accelerators. Developers can provision and manage GPU instances for AI and machine learning workloads through API calls. The service supports on-demand compute for training, fine-tuning, and batch inference jobs, with usage-based billing and enterprise features such as team collaboration and usage tracking. Instances can be accessed via SSH, Jupyter notebooks, or Visual Studio Code, and support quick-deploy templates for popular frameworks including PyTorch, TensorFlow, ComfyUI, n8n, and Langflow.
  version: 1.0.0
  contact:
    name: Qubrid AI Support
    url: https://www.qubrid.com/contact
  termsOfService: https://www.qubrid.com/terms-of-service
servers:
- url: https://platform.qubrid.com/api/v1
  description: Qubrid AI Compute Production Server
security:
- bearerAuth: []
tags:
- name: Chat Completions
  description: Generate chat-based completions using open-source large language models hosted on NVIDIA GPU infrastructure. Compatible with the OpenAI chat completions request and response format.
paths:
  /chat/completions:
    post:
      operationId: createChatCompletion
      summary: Create a chat completion
      description: Generates a model response for the given chat conversation. This endpoint is compatible with the OpenAI chat completions format, accepting an array of messages and returning a generated assistant reply. Supports text generation, code generation, and vision-language models available on the Qubrid AI platform.
      tags:
      - Chat Completions
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ChatCompletionRequest'
      responses:
        '200':
          description: Successfully generated a chat completion response.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ChatCompletionResponse'
        '400':
          description: The request was malformed or contained invalid parameters.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '401':
          description: Authentication failed due to a missing or invalid bearer token.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '404':
          description: The specified model was not found or is not available.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '429':
          description: Rate limit exceeded. Too many requests were sent in a given time period.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
        '500':
          description: An internal server error occurred during inference.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ErrorResponse'
components:
  schemas:
    ChatCompletionResponse:
      type: object
      properties:
        id:
          type: string
          description: A unique identifier for the chat completion.
        object:
          type: string
          enum:
          - chat.completion
          description: The object type, always chat.completion.
        created:
          type: integer
          description: The Unix timestamp in seconds of when the chat completion was created.
        model:
          type: string
          description: The model used for the chat completion.
        choices:
          type: array
          description: A list of chat completion choices. Can be more than one if n is greater than 1.
          items:
            $ref: '#/components/schemas/ChatCompletionChoice'
        usage:
          $ref: '#/components/schemas/Usage'
    ErrorResponse:
      type: object
      properties:
        error:
          type: object
          properties:
            message:
              type: string
              description: A human-readable error message describing what went wrong.
            type:
              type: string
              description: The type of error that occurred.
            code:
              type: string
              description: A machine-readable error code.
    Usage:
      type: object
      properties:
        prompt_tokens:
          type: integer
          description: The number of tokens in the prompt.
        completion_tokens:
          type: integer
          description: The number of tokens in the generated completion.
        total_tokens:
          type: integer
          description: The total number of tokens used in the request (prompt plus completion).
    ChatCompletionRequest:
      type: object
      required:
      - model
      - messages
      properties:
        model:
          type: string
          description: The identifier of the model to use for generating the chat completion. Must be one of the models available on the Qubrid AI platform.
          example: deepseek-ai/DeepSeek-R1-Distill-Llama-70B
        messages:
          type: array
          description: A list of messages comprising the conversation so far. Each message has a role (system, user, or assistant) and content.
          items:
            $ref: '#/components/schemas/ChatMessage'
          minItems: 1
        temperature:
          type: number
          description: Sampling temperature between 0 and 2. Higher values like 0.8 make the output more random, while lower values like 0.2 make it more focused and deterministic.
          minimum: 0
          maximum: 2
          default: 1.0
        top_p:
          type: number
          description: Nucleus sampling parameter. The model considers the results of the tokens with top_p probability mass. A value of 0.1 means only the tokens comprising the top 10% probability mass are considered.
          minimum: 0
          maximum: 1
          default: 1.0
        n:
          type: integer
          description: How many chat completion choices to generate for each input message.
          minimum: 1
          default: 1
        max_tokens:
          type: integer
          description: The maximum number of tokens to generate in the chat completion. The total length of input tokens and generated tokens is limited by the model's context length.
          minimum: 1
        stream:
          type: boolean
          description: 'If true, partial message deltas will be sent as server-sent events as they become available, with the stream terminated by a data: [DONE] message.'
          default: false
        stop:
          oneOf:
          - type: string
          - type: array
            items:
              type: string
            maxItems: 4
          description: Up to 4 sequences where the API will stop generating further tokens.
        presence_penalty:
          type: number
          description: Penalizes new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
          minimum: -2
          maximum: 2
          default: 0
        frequency_penalty:
          type: number
          description: Penalizes new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
          minimum: -2
          maximum: 2
          default: 0
    ChatCompletionChoice:
      type: object
      properties:
        index:
          type: integer
          description: The index of the choice in the list of choices.
        message:
          $ref: '#/components/schemas/ChatMessage'
        finish_reason:
          type: string
          enum:
          - stop
          - length
          - content_filter
          description: The reason the model stopped generating tokens. stop means the model hit a natural stop point or a provided stop sequence, length means the maximum number of tokens was reached, and content_filter means content was omitted due to a filter.
    ChatMessage:
      type: object
      required:
      - role
      - content
      properties:
        role:
          type: string
          enum:
          - system
          - user
          - assistant
          description: The role of the message author. Use system for setting the assistant's behavior, user for the human's input, and assistant for previously generated responses.
        content:
          oneOf:
          - type: string
          - type: array
            items:
              $ref: '#/components/schemas/ContentPart'
          description: The content of the message. Can be a string for text-only messages, or an array of content parts for multimodal messages that include images.
    ContentPart:
      type: object
      required:
      - type
      properties:
        type:
          type: string
          enum:
          - text
          - image_url
          description: The type of content part. Use text for text content and image_url for image content in vision-language model requests.
        text:
          type: string
          description: The text content, used when the type is text.
        image_url:
          type: object
          description: The image URL object, used when the type is image_url.
          properties:
            url:
              type: string
              format: uri
              description: The URL of the image to include in the message.
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer
      bearerFormat: QUBRID_API_KEY
      description: Qubrid AI API key passed as a bearer token in the Authorization header. Obtain your API key from the Qubrid AI platform dashboard at https://platform.qubrid.com.
externalDocs:
  description: Qubrid AI Documentation
  url: https://docs.platform.qubrid.com