Hi Felipe :)\u000a\u000aI know we have 300,000 GPT\u002D3.5 token per device per month, and 1.5 million Mixtral tokens per device per month. My question is, how do we switch models? I can see the model selector on the top of the window, and the usage (percentage) of tokens on the bottom right corner... but that is only displaying GPT3.5 tokens left, even if I switch the model Mixtral, the tokens being used are from the GPT 3.5 model. \u000a\u000aHow do I effectively switch to Mixtral and use Mixtral tokens????\u000a\u000aThanks!

Question

Hi Felipe :)\u000a\u000aI know we have 300,000 GPT\u002D3.5  token per device per month, and 1.5 million Mixtral tokens per device per month. My question is, how do we switch models? I can see the model selector on the top of the window, and the usage (percentage) of tokens on the bottom right corner... but that is only displaying GPT3.5 tokens left, even if I switch the model Mixtral, the tokens being used are from the GPT 3.5 model. \u000a\u000aHow do I effectively switch to Mixtral and use Mixtral tokens????\u000a\u000aThanks!

Felipe_TriploAI · Answer

Hello there!\u000a\u000aSimply choose the model you want to use from the selector. The indicator will still display 300,000 tokens even if you select Mixtral, but each Mixtral token counts as 1/5 of a regular GPT 3.5 token. This means the base remains the same at \u0022300,000,\u0022 but every time you use Mixtral, it will cost you 1/5 of the tokens.\u000a\u000aFor example, if you prompt 1,000 tokens on GPT 3.5, it will deduct 1,000 tokens from the allowance. If you prompt 1,000 tokens on Mixtral, it will deduct 200 tokens. In this way, if you exclusively use Mixtral, you will be able to prompt 1,500,000 tokens (300,000 x 5).\u000a\u000aIf you have any further questions, please send us a message at support@triplo.ai. It\u0027s easier to track support calls there, as our campaign on AppSumo is now over.\u000a\u000aTake care,\u000aFelipe

Triplo AI

Related questions