OpenAI offers several models with different capabilities and price points. This guide helps you choose the right one for your Metorial integration.

Available Models

GPT-4o (Recommended)

• Best for: Most use cases, balanced performance and cost
• Strengths: Fast, multimodal, excellent at tool calling
• Model name: gpt-4o

GPT-4o Mini

• Best for: High-volume, cost-sensitive applications
• Strengths: Very affordable, still capable with tools
• Model name: gpt-4o-mini

GPT-4 Turbo

• Best for: Complex reasoning tasks
• Strengths: Most capable reasoning, larger context window
• Model name: gpt-4-turbo

GPT-3.5 Turbo

• Best for: Simple tasks, budget-conscious projects
• Strengths: Lowest cost, fast responses
• Model name: gpt-3.5-turbo

Specifying a Model

const response = await openai.chat.completions.create({
  model: 'gpt-4o', // Change this to your preferred model
  messages,
  tools: session.tools
});

Recommendations

• Start with GPT-4o - it offers the best balance for most applications
• Use GPT-4o Mini for high-volume chatbots or simple queries
• Choose GPT-4 Turbo when you need advanced reasoning or large context windows
• Consider costs - monitor your usage and adjust based on your budget

Tool Calling Performance

All modern OpenAI models support function calling, but GPT-4o and GPT-4 Turbo tend to make better decisions about when and how to use tools. GPT-3.5 Turbo can work but may be less reliable for complex tool interactions.