Skip to main content

​
Overview

The OpenAI provider gives you access to OpenAI’s models including GPT-5, GPT-4, o-series reasoning models, embeddings, and DALL-E image generation.

​
Installation

npm install @core-ai/openai

​
createOpenAI()

Create an OpenAI provider instance. 
import { createOpenAI } from '@core-ai/openai';

const openai = createOpenAI({
  apiKey: process.env.OPENAI_API_KEY,
  baseURL: 'https://api.openai.com/v1' // optional
});

​
Options

​
apiKey
string
Your OpenAI API key. Defaults to OPENAI_API_KEY environment variable.
​
baseURL
string
Custom base URL for API requests. Useful for proxies or OpenAI-compatible APIs.
​
client
OpenAI
Provide your own configured OpenAI client instance.

​
Returns

OpenAIProvider - Provider instance with methods to create models.

​
Provider Methods

​
chatModel()

Create a chat model instance. 
const model = openai.chatModel('gpt-5.2');
​
modelId
string
required
Model identifier. See Supported Models below.

​
embeddingModel()

Create an embedding model instance. 
const embeddings = openai.embeddingModel('text-embedding-3-large');
​
modelId
string
required
Embedding model identifier (e.g., text-embedding-3-small, text-embedding-3-large).

​
imageModel()

Create an image generation model instance. 
const imageGen = openai.imageModel('dall-e-3');
​
modelId
string
required
Image model identifier (e.g., dall-e-2, dall-e-3).

​
Supported Models

​
Chat Models

  • gpt-5.2 - Latest flagship model with max reasoning effort
  • gpt-5.2-codex - Optimized for code generation
  • gpt-5.2-pro - Enhanced reasoning capabilities
  • gpt-5.1 - Previous generation flagship
  • gpt-5 - Balanced performance and cost
  • gpt-5-mini - Fast and efficient
  • gpt-5-nano - Lightweight model
  • o4-mini - Latest compact reasoning model
  • o3 - Advanced reasoning capabilities
  • o3-mini - Efficient reasoning model
  • o1 - First-generation reasoning model
  • o1-mini - Compact reasoning model (no effort control)
  • gpt-4 - Previous flagship model
  • gpt-4-turbo - Faster GPT-4 variant
  • gpt-4o - Optimized GPT-4

​
Embedding Models

  • text-embedding-3-large - 3072 dimensions, highest quality
  • text-embedding-3-small - 1536 dimensions, faster and cheaper
  • text-embedding-ada-002 - Legacy embedding model

​
Image Models

  • dall-e-3 - Latest image generation model
  • dall-e-2 - Previous generation

​
Capabilities

FeatureSupport
Chat Completion✓
Streaming✓
Function Calling✓
Vision✓
Reasoning Effort✓ (model-dependent)
Embeddings✓
Image Generation✓

​
Examples

​
Basic Chat

import { createOpenAI } from '@core-ai/openai';
import { generateText } from '@core-ai/core-ai';

const openai = createOpenAI({
  apiKey: process.env.OPENAI_API_KEY
});

const result = await generateText({
  model: openai.chatModel('gpt-5.2'),
  prompt: 'Explain quantum computing in simple terms'
});

console.log(result.text);

​
Reasoning with Effort Control

const result = await generateText({
  model: openai.chatModel('gpt-5.2'),
  prompt: 'Solve this complex mathematical proof...',
  reasoning: {
    effort: 'max' // 'minimal' | 'low' | 'medium' | 'high' | 'max'
  }
});

​
Embeddings

import { embed } from '@core-ai/core-ai';

const result = await embed({
  model: openai.embeddingModel('text-embedding-3-large'),
  value: 'Search query text'
});

console.log(result.embedding); // number[]

​
Image Generation

import { generateImage } from '@core-ai/core-ai';

const result = await generateImage({
  model: openai.imageModel('dall-e-3'),
  prompt: 'A futuristic city at sunset',
  size: '1024x1024'
});

console.log(result.images);

​
Custom Base URL (OpenAI-compatible APIs)

const openai = createOpenAI({
  apiKey: 'your-api-key',
  baseURL: 'https://your-proxy.com/v1'
});

​
Reasoning Effort

OpenAI models support different reasoning effort levels:
  • minimal - Fastest, least reasoning
  • low - Quick responses with basic reasoning
  • medium - Balanced reasoning and speed
  • high - Deep reasoning for complex problems
  • max - Maximum reasoning effort (GPT-5.2 series only)
Some models restrict sampling parameters when using reasoning effort. The SDK handles this automatically.

​
Error Handling

import { APIError } from '@core-ai/core-ai';

try {
  const result = await generateText({
    model: openai.chatModel('gpt-5.2'),
    prompt: 'Hello!'
  });
} catch (error) {
  if (error instanceof APIError) {
    console.error('OpenAI API error:', error.message);
    console.error('Status:', error.statusCode);
  }
}

Anthropic Provider

Claude models with extended thinking

Google GenAI Provider

Gemini models with multimodal capabilities

Core AI Functions

Learn about generate, stream, and more

Build docs developers (and LLMs) love

Get started for freeTalk to us