Token Price Comparison of AI Models: Why Result Cost Matters

Thibaut from OpenAI came for Opus.

First, he asked why companies still use Opus instead of Sol and whether price really matters to them.

Then he explained his position in more detail.

"One token from OpenAI does not equal one token from another model."

We compare AI prices in dollars per million tokens as if a token is a standardized unit like a gram or a kilowatt-hour. But that’s not the case.

Thibaut gives a simple example with pizza. One pizza is cut into 8 slices at $2 each, another into 16 slices at $1.25 each. In the second case, a slice is cheaper, but the whole pizza costs more.

The same goes for models. In a small test, GPT-5.6 Sol used 766 tokens for the same text that Claude Opus 5 needed 1,170 tokens for.

That’s 34.5% fewer tokens for the same result.

Therefore, a lower price per million tokens does not necessarily mean that the model is actually cheaper. You need to consider the cost of a successful outcome, not just the price of a single token.