I don’t know how I did but I know one thing: in 3 days only I used my 1,000,000,000 tokens limit for the model mistral/devstral-2512 on my Mistral AI API account!
To do that, I used Mistral Vibe. Maybe this is the cause because of the very big context (the environment, etc…) and the very big generated answers (actions in the environment, explanations on the reasoning, etc…).
Happily, my limit reinitializes on March 1 which is Sunday!
Top comments (0)