Not showing the uncached token count prominently makes the total token usage seem much higher than it is as cached tokens are very cheap.
Gemini 3.0 makes a large # of small requests that much better utilize cached tokens which makes the total token count very misleading.
We should instead show the number of non-cached tokens as the input tokens as that is the meaningful stat for most users.
We should likely indicate the # of cached tokens at least for API Key users but need to make sure we don't do it in a way that causes people to be confused about cost of overall efficiency.
Not showing the uncached token count prominently makes the total token usage seem much higher than it is as cached tokens are very cheap.
Gemini 3.0 makes a large # of small requests that much better utilize cached tokens which makes the total token count very misleading.
We should instead show the number of non-cached tokens as the input tokens as that is the meaningful stat for most users.
We should likely indicate the # of cached tokens at least for API Key users but need to make sure we don't do it in a way that causes people to be confused about cost of overall efficiency.