Are the thinking tokens counted in the output price for 2.5 Flash?

Hi,

I’m looking for some clarification regarding the pricing for Gemini 2.5 Flash, specifically concerning “thinking” tokens.
According to the Gemini API documentation, the output price per 1 million tokens is:

  • Non-thinking: $0.60
  • Thinking: $3.50

My question is: When thinking is on, are thinking tokens counted as output tokens, similar to how they are for Gemini 2.5 Pro?

Thanks for your help!

When thinking is turned on, response pricing is the sum of output tokens and thinking tokens.

https://ai.google.dev/gemini-api/docs/thinking#pricing