CLI impact on tokens?

[CXify]

So, I've built a competitive intelligence benchmarking tool with base44. I've built 4-5 prototypes to figure out what I can/can't do with basic LLM prompting, etc. My prototypes started mixing canonical data from different db schemas in base as when I built new versions I had asked to iterate on old ones and it just went and created all new ones from scratch! #learning. This was frustrating, but I've marched on! This is why I have 4-5 prototypes and not 2 iterated ones. The db was such a mess that I decided to extract the next iteration of the pipeline process into an n8n workflow with supabase backend. So. Much. Learning. Now I have some in base and some in supabase/n8n - which is consistent and reliable.

I saw on another thread CLI is available for all tiers, but can you clarify how it will impact tokens? I had tried out Agents - i built a few employees to do some work with, but my tokens burned in an hour. I didn't really understand that would happen and I had to upgrade to continue. I now see this as the cost of learning to build what I wanted as a novice. Now that I'm more efficient about when to use them, when not to, I'm more confident in how to use them and less annoyed when chatAI fails to fix something or repeats the same mistakes at cost. I've learned to edit code and schema myself and when you launched git repo, I spun up cursor with claude code and started building my apps more efficiently - that's been awesome.

Now, I'm using Claude as a my thought-partner/planning; cursor with CC for backend and code and Antigravity for UX/design and here you come along with direct base44 CLI, so now i'll spin that up in my next sprint.

How do I think about the tokens will be burned on this in the future? will the 100 credits basically be burned by this trial? Whatabout moving forward? Will I be able build everything in CLI at no token cost? If it spins up code skils/agents, will they run at a high token rate that's hard to comprehend? How will you burn tokens - will be it be processing with AGENTs or ihternal LLM use?
How will my token use be affected when I launch to real customers?

Thanks for any clarity so I can plan my apps and scaling them more effectively.

[tuvit]

Hey, thanks for sharing! To clarify on tokens specifically: the Base44 CLI itself doesn't consume any tokens - it's a project management and deployment tool. You only use Base44 tokens when you explicitly invoke AI features like agent skills or AI-powered backend functions in your app. So the CLI won't burn through your token allocation, but tokens are consumed when AI capabilities are actively used. Hope this helps with your planning :)