Adding the used tokens count is a very nice update. My question is:
1: Is the 90000 my Copilot limit? Is it getting that value directly from Copilot?
2: What are Copilot Pro users seeing as a limit?
Bonus Praise: I tried out the Sample ToDo App prompt. Very impressive. 
Also 90000. It's difficult to know how the limits and usage are calculated. What about 'completions':
?
References to completions appear in Wappler's terminal:
Anybody understand how this works?
The limits differ per model. For the Claude 3.5 is 90k tokens the limit for a single chat (request) so the maximum you can send in one time.
As you can see most of it is prefilled by us with predefined knowledge and rules to follow.
Then additional context is added about your selection and your question and that is send as a single request.
Indeed with the free plan you can have only 50 of those chat requests per month.
Code completions are the completions when you edit directly in the code editor, it will then give you code suggestions. With the free plan you get a maximum of 2000 suggestions per month.
For the chat, which is used for the AI Assistant you can do 50 messages per month. There is no token limits.
The 90000 token limit is the maximum that the selected model accepts for the prompt, the output is limited to something like 4096 or 8192 tokens. The full context window of the models are often even bigger.
Sometimes when I am working in code view I get code suggestions in grey text. Is that from the AI assistant? I notice sometimes it is pretty clever. If I wanted to use that suggested code how would I do that?
Thanks again, I have so much to learn about this AI stuff.
Yes you can then press the tab key to accept it
