
How much are you spending on “Vibe Coding”? If you’re creating smaller projects, to be honest, it’s probably not a large amount, but here’s what we gathered from Gemini (USD):
User | Range | Tools |
Individual hobbyist | $0 – $25/month | ChatGPT, Copilot, Gemini |
Professional Developer | $200 – $1,500+ / year | Cursor Pro |
Small to Medium-sized Team | $2,500 – $15,000+ / year | Github Copilot Business |
Large Enterprise | $25,000 – $1M+ / year | Cursor Enterprise, Gemini for Workspace Enterprise. Model Training. |
How long should my prompts run?
If you’ve been vibe coding – or preferably prompt engineering – how long do you let the AI run for? We’ve heard stories of 5-10 minutes. Note that the time will increase the bigger your projects are. This will cost you in the long run, so how do we reduce the cost? Make it run less.
Optimally, in the beginning, you’ll want prompts that run for 1 minute or less at a time; we’re talking about creating 1 skeleton page or function block at a time.
Reducing Prompt Time
For the average developer, creating 1 template or function at a time is self-explanatory, but how do we approach this if you don’t know what a function is? The natural approach here in terms of measurement is: a task or scenario. For example, if you wanted something to happen when a button is clicked, describe that button and describe what you want to happen, as small a task as you can.
When I click the add button, i want a new page modal to pop up.
That’s it. Avoid going into more detail, like what that page does or what it’s going to do. This will reduce your costs.
I know 10-30 dollars doesn’t seem like much per month, but we should be saving where we can. The key is to execute smaller tasks, which our guide on prompt covers.
In It For The Long Run
For larger projects, you’ll want to approach this at scale. Meaning documentation, see our enterprise recommendation here.
Remember, vibe coding is a tool, not a shortcut, and it’s up to us to use it properly. Especially developers, this is no excuse, as you’ll be leading the future of AI prompt engineering.
0 Comments