AI Streaming API

An AI/LLM API example demonstrating streaming responses with Server-Sent Events.

Features Demonstrated

SSE streaming configuration
Dual response types (streaming and non-streaming)
Sentinel event filtering
Long timeout configuration
Streaming SDK usage patterns

OpenAPI Specification

openapi: 3.1.0
info:
  title: AI Completions API
  version: 1.0.0

servers:
  - url: https://api.ai.example.com/v1

security:
  - bearerAuth: []

paths:
  /chat/completions:
    post:
      operationId: createChatCompletion
      summary: Create a chat completion
      description: |
        Generate a chat completion. When `stream: true`, returns
        Server-Sent Events stream of completion chunks.
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ChatCompletionRequest'
      responses:
        "200":
          description: Chat completion
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ChatCompletion'
            text/event-stream:
              schema:
                $ref: '#/components/schemas/ChatCompletionChunk'

  /embeddings:
    post:
      operationId: createEmbedding
      summary: Create embeddings
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/EmbeddingRequest'
      responses:
        "200":
          description: Embeddings
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/EmbeddingResponse'

  /models:
    get:
      operationId: listModels
      summary: List available models
      responses:
        "200":
          description: Model list
          content:
            application/json:
              schema:
                type: object
                properties:
                  data:
                    type: array
                    items:
                      $ref: '#/components/schemas/Model'

components:
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer

  schemas:
    ChatMessage:
      type: object
      required:
        - role
        - content
      properties:
        role:
          type: string
          enum: [system, user, assistant]
        content:
          type: string
        name:
          type: string

    ChatCompletionRequest:
      type: object
      required:
        - model
        - messages
      properties:
        model:
          type: string
          example: "gpt-4"
        messages:
          type: array
          items:
            $ref: '#/components/schemas/ChatMessage'
          minItems: 1
        temperature:
          type: number
          minimum: 0
          maximum: 2
          default: 1
        max_tokens:
          type: integer
          minimum: 1
        stream:
          type: boolean
          default: false

    ChatCompletion:
      type: object
      required:
        - id
        - model
        - choices
      properties:
        id:
          type: string
        model:
          type: string
        choices:
          type: array
          items:
            type: object
            properties:
              index:
                type: integer
              message:
                $ref: '#/components/schemas/ChatMessage'
              finish_reason:
                type: string
                enum: [stop, length, content_filter]
        usage:
          type: object
          properties:
            prompt_tokens:
              type: integer
            completion_tokens:
              type: integer
            total_tokens:
              type: integer

    ChatCompletionChunk:
      type: object
      required:
        - id
        - model
        - choices
      properties:
        id:
          type: string
        model:
          type: string
        choices:
          type: array
          items:
            type: object
            properties:
              index:
                type: integer
              delta:
                type: object
                properties:
                  role:
                    type: string
                  content:
                    type: string
              finish_reason:
                type: string
                enum: [stop, length, content_filter]
                nullable: true

    EmbeddingRequest:
      type: object
      required:
        - model
        - input
      properties:
        model:
          type: string
        input:
          oneOf:
            - type: string
            - type: array
              items:
                type: string

    EmbeddingResponse:
      type: object
      properties:
        data:
          type: array
          items:
            type: object
            properties:
              index:
                type: integer
              embedding:
                type: array
                items:
                  type: number
        usage:
          type: object
          properties:
            prompt_tokens:
              type: integer
            total_tokens:
              type: integer

    Model:
      type: object
      properties:
        id:
          type: string
        created:
          type: integer
        owned_by:
          type: string

Soxom Configuration

version: "1.0"

sdk:
  name: ai-sdk
  version: 1.0.0
  description: SDK for AI Completions API

spec:
  path: ./openapi.yaml

targets:
  - typescript
  - python

resources:
  $client:
    methods:
      listModels: get /models

  chat:
    subresources:
      completions:
        models:
          - ChatCompletion
          - ChatCompletionChunk
          - ChatMessage
        methods:
          create: post /chat/completions

  embeddings:
    methods:
      create: post /embeddings

authentication:
  default: bearer
  env_vars:
    bearer:
      token: AI_API_KEY

# Streaming configuration
streaming:
  sse:
    enabled: true
    sentinel_events:
      - "[DONE]"

# Longer timeouts for AI operations
timeouts:
  default_ms: 120000    # 2 minutes
  operations:
    chat_completions_create:
      default_ms: 300000  # 5 minutes for long completions

# Retry on rate limits
retries:
  enabled: true
  max_attempts: 3
  retry_on:
    - 429
    - 500
    - 502
    - 503

Generated SDK Usage

import { AIClient } from 'ai-sdk';

const client = new AIClient({
  token: process.env.AI_API_KEY
});

// Non-streaming (default)
const completion = await client.chat.completions.create({
  model: "gpt-4",
  messages: [
    { role: "system", content: "You are a helpful assistant." },
    { role: "user", content: "Explain quantum computing in simple terms." }
  ],
  temperature: 0.7,
  max_tokens: 500
});

console.log(completion.choices[0].message.content);
console.log(`Tokens used: ${completion.usage.total_tokens}`);

from ai_sdk import AIClient

client = AIClient(token=os.environ["AI_API_KEY"])

# Non-streaming (default)
completion = client.chat.completions.create(
    model="gpt-4",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Explain quantum computing in simple terms."}
    ],
    temperature=0.7,
    max_tokens=500
)

print(completion.choices[0].message.content)
print(f"Tokens used: {completion.usage.total_tokens}")

Streaming Completion

TypeScript
Python

// Enable streaming
const stream = await client.chat.completions.create({
  model: "gpt-4",
  messages: [
    { role: "user", content: "Write a short story about a robot." }
  ],
  stream: true
});

// Process chunks as they arrive
for await (const chunk of stream) {
  const content = chunk.choices[0]?.delta?.content;
  if (content) {
    process.stdout.write(content);
  }

  // Check for completion
  if (chunk.choices[0]?.finish_reason === "stop") {
    console.log("\n[Complete]");
  }
}

# Enable streaming
stream = client.chat.completions.create(
    model="gpt-4",
    messages=[
        {"role": "user", "content": "Write a short story about a robot."}
    ],
    stream=True
)

# Process chunks as they arrive
for chunk in stream:
    content = chunk.choices[0].delta.content
    if content:
        print(content, end="", flush=True)

    # Check for completion
    if chunk.choices[0].finish_reason == "stop":
        print("\n[Complete]")

Collecting Streamed Response

TypeScript
Python

const stream = await client.chat.completions.create({
  model: "gpt-4",
  messages: [{ role: "user", content: "Hello!" }],
  stream: true
});

// Collect all content
let fullContent = "";
for await (const chunk of stream) {
  fullContent += chunk.choices[0]?.delta?.content ?? "";
}

console.log("Full response:", fullContent);

stream = client.chat.completions.create(
    model="gpt-4",
    messages=[{"role": "user", "content": "Hello!"}],
    stream=True
)

# Collect all content
full_content = ""
for chunk in stream:
    full_content += chunk.choices[0].delta.content or ""

print("Full response:", full_content)

Embeddings

TypeScript
Python

// Single text
const embedding = await client.embeddings.create({
  model: "text-embedding-ada-002",
  input: "Hello, world!"
});
console.log(`Dimensions: ${embedding.data[0].embedding.length}`);

// Multiple texts
const embeddings = await client.embeddings.create({
  model: "text-embedding-ada-002",
  input: ["Hello", "World", "AI"]
});
console.log(`Generated ${embeddings.data.length} embeddings`);

# Single text
embedding = client.embeddings.create(
    model="text-embedding-ada-002",
    input="Hello, world!"
)
print(f"Dimensions: {len(embedding.data[0].embedding)}")

# Multiple texts
embeddings = client.embeddings.create(
    model="text-embedding-ada-002",
    input=["Hello", "World", "AI"]
)
print(f"Generated {len(embeddings.data)} embeddings")

List Models

TypeScript
Python

const models = await client.listModels();
for (const model of models.data) {
  console.log(`${model.id} (${model.owned_by})`);
}

models = client.list_models()
for model in models.data:
    print(f"{model.id} ({model.owned_by})")