environment variables in the new dashboard
# šŸ¤help
p
I am trying the new dashboard and I can't find where are the environment variables here... could someone help me please?
f
@plain-lifeguard-16679šŸ‘€
For now you can find them in your bot
p
alright thank you! @fresh-fireman-491 will the environment variables tab be back in the dashboard? that was useful because it was faster
f
I only think that @plain-lifeguard-16679 could give us an answer about that
p
hahah where are you chris??
since i'm here, i'd have some questions about the ai spend calculator. can i ask you or do i have to write to the sales team?
f
I think it's still early in Canada so they have probably not clocked into work yet.
Probably better to create a new help post about the new question
p
alright, thank you very much decay<3
f
You are very welcome
p
Hey sorry I'll add this back in the new dashboard soon! in the mean time you can switch back to the old one just to configure them šŸ™‚
I can try to help but I can't garantee anything šŸ˜…
p
it's alright, i've read some other help questions from other users meanwhile and I think you'd answer the same to me
its about that ai spend estimated price. the two prices that the calculator gives are a bit too far away from each other, so i wanted to know more about how it's calculated
f
AI service providers charge for AI tokens used to do things like answer questions and generate text. AI Spend is charged at cost from these service providers, no hidden fees or extra costs. You can find the price of third party AI tokens by directly visiting the pricing page of the tokens' model and getting the information from the model provider. Since Botpress does not mark-up AI tokens' cost, the price charged will reflect the cost incurred by Botpress for that respective token at that time. For OpenAI's GPT 3.5 Turbo and GPT 4 model pricing, please refer to the AI tokens' cost in their official pricing page. For Bing Search API pricing, please refer to the cost in their official pricing page. Any action that requires AI capabilities will consume AI tokens, applicable both in conversation with the bot and in the Studio while building and testing in the Emulator. Here are a few examples: ā€ ā€¢ Knowledge Base ingestion & answering ā€¢ Personality Agent rewrites ā€¢ Translator Agent language identification & message translation ā€¢ Summary Agent conversation summarization ā€¢ AI Task, AI Transition, and AI Generate Text cards ā€¢ Execute Code card's generative AI code assist ā€¢ Capture cards ā€¢ Making Table columns searchable ā€¢ Transforming user sentences to a vector
p
so a message can cost me 0.001 dollars but could also cost me like 0.1
depending on how many tokens the incoming and the outgoing messages have?
f
Yes
And which model you use
GPT-4 is around 30x as expensive as GPT-3.5
p
alright thank you very much for your time
f
You are very welcome