@alihamza79 @david_whisper
Sorry for the late reply, I had to “clock-in” first before I shilled Cursor some more.
In all honesty, I think you guys may not know how to use Cursor frugally or are straight up ignoring how much requests cost. They have larger plans than Pro for a reason, so you can’t use Pro like its Ultra and then be upset when it runs out in 3 hours.
How about after you make a request, you check how much it cost, so you get an idea of how to use the tool? Every time someone says “omg I used sonnet 4.5 and it only lasted 3 hours on my $20 plan”, I wonder how dumb they have to be to not notice that its eating up their quota fast. Its like you want your balance to run empty so you can complain. Gpt-5 would at least get you 3x+ more requests than sonnet 4.5. But of course, I am some shill because I know how to use the tool frugally when I say use the FREE models like gpt-5-mini and grok-code-fast-1.
You have to admit you guys don’t know how to use this tool. You either
- don’t pay attention to the cost per request
- don’t choose the least expensive (sometimes free) model based on the requirements of the request
- you are using Cursor with massive codebases, without specifying clear context
- you are not making new chats when context is getting too big
- your requests are asking to do too much at a time, making them very expensive and requiring models that can handle huge requests instead of using cheaper models and focusing on a feature/bug at a time
- you are vibe coding and spamming requests in a queue
Is this how manage your money in any other area of your life? Do you shop at the grocery store by buying the most expensive option for every item on your list? No you probably try to determine what is the best value based on your personal needs/requirements for each item. It’s $20 man, you can’t expect it to provide $$$ of requests. The requests cost us, I mean Cursor, money:
Gpt-5 is like 60% cheaper.
And the free models are infinitely cheaper, but even if they did count towards your usage, look how much more your $20 would go.
There are plenty of people on this forum praising these models. These are no doubt the best way to get the most out of Cursor if you are on a $20/mo budget (which I would probably claim means you are not a professional, which is okay, just need to keep things in perspective). Hiring another developer that could do what these free models do inside Cursor, would cost you well over $20/mo.
All of these cheaper models are very effective at certain requests. However, I use sonnet 4.5 when the others fail or when I know from experience that the others will fail. I also use sonnet 4.5 (or gpt-5) when I am adding a new feature to create the plan and the initial start, then I have the cheaper models sort of fill in the code and make minor adjustments. I jump out of using sonnet 4.5 as soon as I can.
Are you guys professionals? Like do you actually get paid for developing and have been doing this for 5+ years? Because any legit developer would see the value in this $20 plan, especially compared to the alternatives, and would not be surprised that using the most expensive model was uhh expensive.
I actually pay $16/mo (annual which makes it even cheaper, but that is before my “employee discount” 
) and get a pretty good product. If you don’t get more than $16/mo of value, you’re not a programmer and you are a cheapskate who has no idea what value is. Go to some cheaper alternative like Trae or Kiro or whatever. Cursor doesn’t need your complaining because you want more stuff for free.
Reasons why Cursor is good (because you claim I never give reasons):
- integrate into the IDE intimately
- have (unlimited) tab completions
- automatically create restore checkpoints made with each request
- allow inline requests
- show diffs and allow to manually choose which changes to accept
- Context and Awareness
- automatically gather context with open tabs and looks through your project folder
- monitor clipboard/selection and which files and specific sections of code you are just simply looking at, let alone interacting with
- allow copying a section of code into the request to give it very clear context (code section and file)
- allow dragging tabs, files, sections of code, and pasting screenshots all into the request context
- duplicate a chat to create a sort of branch that you can return to
- easily switch between models (some free)
- Clear privacy policy for not training on code
Reasons why Cursor is bad (because you claim I never critique):
- broken trust by changing pricing plans multiple times
- updates that break previous workflow or forcing new features as default (however it is good to see they actively listen to the forum for feedback and are making improvements to the software)
- more expensive than some alternatives
It is possible someone just likes Cursor and is annoyed when people just spam that “Cursor suc ks” without giving clear explanations or just begging for Cursor to be cheaper. It’s a products, and a good one, and I am surprised I can get away with only paying $16/mo. Go use Trae, Kiro, or Windsurf if price is the primary factor. (yes, I am a “Cursor PR guy” recommending you to go use a different product because we don’t need your business.)
“Phew, shilling is hard work. Time to take my 3 hour paid lunch break while I get my feet massaged by Cursor’s in-house masseuse. God I love Cursor paying me to provide reasonable advice and explanations to their decent product. It’s so easy, I don’t even have to make anything up.”
Why do I make so many posts? Because its something to do while requests run or while waiting for my code to recompile.