OpenAI Quietly Raises API Prices for Older GPT Models

What Happened

OpenAI has raised the prices of several older API models, including some GPT-3.5 variants. The increases range from 20 to 50 percent depending on the model tier. At the same time, OpenAI has kept pricing stable or reduced costs on its newer models like GPT-4o.

Why It Matters

If your business or product was built on cheaper legacy OpenAI models to keep costs down, your monthly bill may have already gone up. This is a common pattern in the AI industry as providers push users toward newer, more capable tools.

The good news is that newer models are often faster and better than the ones being phased out. However, switching requires testing to make sure your prompts and workflows still produce the results you expect.

What To Do

Log in to your OpenAI usage dashboard and check which models your account is calling. If you are running legacy models, compare the cost of switching to GPT-4o mini, which is both cheaper and more capable than GPT-3.5 for most tasks.

Founders building on the API should also consider this a good moment to review whether OpenAI is still the right provider, or whether Gemini, Claude, or open-source alternatives offer better value for your specific use case. Locking into one provider without regular reviews is an easy way to overspend.