Amazon Nova Inference API

openapi: 3.1.0
info:
  title: Amazon Nova on Amazon Bedrock Runtime Async Inference API
  version: '2023-09-30'
  description: 'Invoke Amazon Nova foundation models (Premier, Pro, Lite, Micro, Canvas,

    Reel, Sonic) through the Amazon Bedrock Runtime API. Supports

    request/response and streaming inference using the Converse and

    InvokeModel families of operations, plus asynchronous invocation for

    long-running tasks like Nova Reel video generation.

    All requests are signed with AWS Signature Version 4 (SigV4).

    '
  contact:
    name: Amazon Web Services
    url: https://docs.aws.amazon.com/nova/latest/userguide/what-is-nova.html
servers:
- url: https://bedrock-runtime.{region}.amazonaws.com
  description: Bedrock Runtime regional endpoint
  variables:
    region:
      default: us-east-1
      description: AWS region identifier supporting the desired Nova model.
security:
- SigV4: []
tags:
- name: Inference
  description: Synchronous and streaming inference operations.
paths:
  /model/{modelId}/invoke:
    post:
      tags:
      - Inference
      summary: Invoke a Nova model
      description: 'Synchronously invoke a Nova model identified by its model ID

        (for example `amazon.nova-pro-v1:0`, `amazon.nova-canvas-v1:0`,

        or an inference profile id like `us.amazon.nova-lite-v1:0`).

        '
      operationId: invokeModel
      parameters:
      - in: path
        name: modelId
        required: true
        schema:
          type: string
        description: Nova model identifier or inference profile id.
      - in: header
        name: Content-Type
        required: true
        schema:
          type: string
          enum:
          - application/json
      - in: header
        name: Accept
        schema:
          type: string
          enum:
          - application/json
      requestBody:
        required: true
        content:
          application/json:
            schema:
              type: object
              description: Nova-specific request payload (messages, inferenceConfig, etc.).
      responses:
        '200':
          description: Model output.
          content:
            application/json:
              schema:
                type: object
  /model/{modelId}/invoke-with-response-stream:
    post:
      tags:
      - Inference
      summary: Invoke a Nova model with streaming output
      operationId: invokeModelWithResponseStream
      parameters:
      - in: path
        name: modelId
        required: true
        schema:
          type: string
      requestBody:
        required: true
        content:
          application/json:
            schema:
              type: object
      responses:
        '200':
          description: Event stream of model output chunks.
  /model/{modelId}/converse:
    post:
      tags:
      - Inference
      summary: Converse with a Nova model
      description: Use the unified Bedrock Converse API for chat-style interactions.
      operationId: converse
      parameters:
      - in: path
        name: modelId
        required: true
        schema:
          type: string
      requestBody:
        required: true
        content:
          application/json:
            schema:
              type: object
              properties:
                messages:
                  type: array
                  items:
                    type: object
                system:
                  type: array
                  items:
                    type: object
                inferenceConfig:
                  type: object
                toolConfig:
                  type: object
      responses:
        '200':
          description: Conversation response.
  /model/{modelId}/converse-stream:
    post:
      tags:
      - Inference
      summary: Converse with streaming
      operationId: converseStream
      parameters:
      - in: path
        name: modelId
        required: true
        schema:
          type: string
      requestBody:
        required: true
        content:
          application/json:
            schema:
              type: object
      responses:
        '200':
          description: Event stream of conversation chunks.
components:
  securitySchemes:
    SigV4:
      type: apiKey
      in: header
      name: Authorization
      description: AWS Signature Version 4 signed request.