GPT-4o token cost & pricing

GPT-4o is OpenAI's previous-generation multimodal flagship, still widely used because it handles text, images and audio in a single model with a 128K context window. Many production systems remain on it, so its pricing matters even as newer models take the spotlight.

GPT-4o specifications

What GPT-4o costs in practice

Four common workloads, priced at list rates. Monthly figures assume 30 days.

How many requests does $100 buy?

Assuming a 1,000-token prompt and a 500-token response per request.

How GPT-4o pricing actually works

At $2.50 per million input and $10 per million output, GPT-4o is more expensive on input than GPT-5 while matching it on output — an unusual position that reflects newer input pricing on the GPT-5 line rather than GPT-4o getting cheaper. Cached input is $1.25, a 50% discount rather than the 90% available on GPT-5.

When to use GPT-4o

Choose GPT-4o when you need image or audio understanding and have already validated prompts against it, or when you depend on features that exist only in this generation. For pure text work, GPT-5 and GPT-5 mini are cheaper and stronger.

How to cut your GPT-4o bill

Migration is the biggest saving available: move text-only traffic to GPT-5 mini and reserve GPT-4o for multimodal calls. Because caching is only 50% off here, prompt-prefix reuse matters less than it does on the GPT-5 line.

Other OpenAI models

Cheaper alternatives to GPT-4o

• GPT-5OpenAI10% cheaper
• Gemini 2.5 ProGoogle10% cheaper
• Mistral Large 2Mistral36% cheaper
• Qwen3 MaxAlibaba42% cheaper