openapi: 3.1.0
info:
  title: Amazon Nova on Amazon Bedrock Runtime 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.
tags:
  - name: Inference
    description: Synchronous and streaming inference operations.
  - name: Async
    description: Asynchronous invocation for long-running generation jobs.
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.
  /async-invoke:
    post:
      tags: [Async]
      summary: Start an asynchronous Nova invocation
      description: Used for Nova Reel and other long-running generations.
      operationId: startAsyncInvoke
      requestBody:
        required: true
        content:
          application/json:
            schema:
              type: object
              properties:
                modelId: { type: string }
                modelInput: { type: object }
                outputDataConfig: { type: object }
      responses:
        "200":
          description: Invocation ARN.
  /async-invoke/{invocationArn}:
    parameters:
      - in: path
        name: invocationArn
        required: true
        schema: { type: string }
    get:
      tags: [Async]
      summary: Get async invocation status
      operationId: getAsyncInvoke
      responses:
        "200":
          description: Invocation status.
components:
  securitySchemes:
    SigV4:
      type: apiKey
      in: header
      name: Authorization
      description: AWS Signature Version 4 signed request.
security:
  - SigV4: []