For the new max mode pricing as shown below what exactly does this mean in terms of pricing? It looks like it means it just uses requests rather than pay per use but that raises a few immediate questions.
What happens when a user runs out of fast requests does max mode stop working? Does it become usage based pricing? Does it just use slow requests? Is there a way to opt into usage based pricing earlier to avoid burning through fast requests in 20-50 max requests?
I understand this pricing model is meant to be simpler but I feel there is information lacking on implementation details. I havenât updated to the new version yet as I want to avoid accidentally burning through all of my fast credits immediately.
Any clarification would be greatly appreciated!
P.S. I know the usage based pricing is cost+20% but if the actual price per model could be listed somewhere under usage based pricing thatâd be very helpful.
It does not prevent you from running out of requests. Youâll still burn through your requests in 20-50 max prompts. It allows you to pay for additional requests when you run out instead of switching to free, slow requests. It doesnât say anything about enabling it before you run out of requests(just that you have to enable it when you run out to keep using fast requests), but you can turn it on or off at any time.
The new pricing method is quite expensive in terms of request usage. I lost hundreds of requests for very minimal coding effort â something that wasnât an issue with previous versions.
Considering this, itâs becoming unaffordable, and competitors are gaining an edge because of it.
I hope Cursor will reconsider its approach to usage-based costing.
Agreed, with non-MAX mode the pricing is very predictable and with the right prompts you can do a lot of coding with 500 requests.
For MAX Mode the pricing depends on tokens, so sending too much context unnecessarily would also consume tokens quickly. Some models have caching tokens (cheaper than first submission) for prompts they already received when you continue in chat.
Normal mode: Fixed cost per message, super predictable
MAX mode: Token-based pricing, can get expensive if sending lots of context
The token caching in MAX mode is pretty neat - when the model sees the same content again (like in follow-up messages), it only charges about 10% of the original token cost. But yeah, if youâre working with a tight request budget, sticking to normal mode is your best bet
- What happens when a user runs out of fast requests does max mode stop working? Does it become usage based pricing? Does it just use slow requests?
If a user uses all of their fast requests, they will no longer have access to MAX unless they enable usage-based pricing. Their is no slow pool for MAX mode requests. - Is there a way to opt into usage based pricing earlier to avoid burning through fast requests in 20-50 max requests?
Unfortunately not, however, your end-of-month bill would be the same hypothetically.
Obviously staying out of MAX mode is going to make costs predictable, but that does not address the OPâs (or my) question â for those times when MAX mode is necessary, how can we understand the costs of the various models?
Also I see the only MAX model is sonnet - gemini went away with the last update?
Not sure if that helps but once you use it, you can see in the new dashboard, how many requests/tokens each actual model call consumed that was triggered from a single max mode prompt. It lists many lines with each stating costs of e.g. 0.2 requests or 2.4 and when hovering you see the tokens.
During usage of max mode, Iâve had some cases where it would go into 30 requests or so and from what Iâve seen I suspect that the amount of tool calls also play a role because cursor internally then re-requests the model (with hopefully many cached tokens) after every single one like list me the files in that directory, show me installed packages et cetera.
I think over time we will figure out prompts that will try to avoid the most expensive usages (unless we are willing to pay for it).
Just a follow-up question I have on the matter. For anyone that is just regularly using any of the max models, whether you are paying out of pocket or your employer is paying it, how much do you generally rack up in spending per day or per month?
I have had a nice experience with rare calls to the max models with the previous pricing model, but I am not sure how quickly the expense builds up primarily using a max model with the price changes. Especially since you canât opt into usage-based pricing for a max model specifically since it now burns through fast requests first. Iâd imagine it would force you into usage-based pricing even for regular models certain times of day if the slow queue is bad.
Anyway, if anyone is willing to share their experience with heavy max usage, Iâd love to hear your thoughts!
It would be a game changer to have a little counter which shows how many tokens are being sent in each request (out of the max) and how much the request costed after it finished.
Guys, whatâs going on? 2 hours and 500 requests have flown away. The agent couldnât even launch a test and perform a mid-term one. What kind of pricing policy is this?! Itâs easier to launch a local open model by buying the equipment once and for all.
Frustration with the New Request Pricing Structure â Power Users Penalized?
Hi all,
Iâm writing as a longtime power user of Cursor AI who has come to rely on the platform for daily coding and productivity. I have invested a significant amount of money into Cursor Pro, specifically purchasing large blocks of requests under the previous yearly Pro plan.
However, with the recent shift from request blocks to a straight $0.04 per request pricing model, I feel incredibly frustrated and, frankly, penalized for being an early and heavy supporter.
Hereâs the math that has me upset:
I paid $1,536 for 4,000 requests (thatâs $0.384 per request).
Under the new structure, 4,000 requests would cost only $160 ($0.04 Ă 4,000).
That means anyone who prepaid at the old rate paid nearly ten times more than a new customer would for the exact same usage!
This isnât just a small pricing change â itâs a massive gap that leaves early Pro supporters at a disadvantage.
Iâve optimized my workflow to maximize my request utility, but this new model makes it feel like all the investment in advance blocks was for nothing. If Iâm misunderstanding something, please clarify. Otherwise, I believe Cursor AI should seriously consider refunding, crediting, or otherwise making things right for their most dedicated users who supported the platform under the old structure.
Has anyone else experienced this? Did you pay for a yearly Cursor Pro and now feel shortchanged?
Would love to hear how others are dealing with this change and whether the Cursor team plans to address it.
Hi, Iâm confused by how you paid almost $1600 for 4000 requests.
In our old requests blocks ($20 for 500 requests), our current Pro plan (also $20 for 500 requests) and our standard usage-based pricing ($0.04/requests = $20/500), you should only ever pay $0.04 for each unit of ârequestâ.
Some models and feature have >1 request cost, such as thinking models, super expensive models and so on, but they would be shown in your usage-pricing breakdown and on our models page.
The only exception to this is Max mode, which had the following pricing:
Old: $0.04 per request and $0.04 per tool call
New: Cost is directly correlated to request length (aka API pricing) - long request = higher cost
In this case, while a ârequestâ could cost much more than $0.04, your costs to Cursor would match up with the costs you incur for us with the LLM provider.
Can you provide any further clarify on how youâve reached this figure, and what that is made up from?
Well I am happy to show you my account, because my math and the numbers around it I went over multiple times and unless there is something wrong with my Cursor Pro packages I am not understanding the math is mathing for me.
Now this only represents my Cursor Pro, which I have also paid monthly for the MAX usage prior to the change. I am currently around $2000~+ invested into Cursor usage for the last 8 months.
You can see by this that $192/ea Cursor Pro block of +500 = 4000 for a total of $1536
I think what you are saying is you paid $1,536 for 8 pro accounts for a year, and you think you only got 500 requests so that makes 4,000 requests.
Thatâs per month, you actually have 48,000 requests. Which works out at $0.032 per request.
But that canât be what you mean, is it? As that just sounds a bit simple, did you just forget you have 500 requests per month, not per year?
If so itâs good news, you are better off than you think, if you just had the one account with usage based pricing for all requests over the 500 youâd spend $32 a month more than you are spending now. Personally for a dollar a day Iâd probably choose to just have the one account as itâs just too much hassle managing 8 which I assume you need to log in and out of? But yea, you are saving money so no need to feel upset at being ripped off - you arenât being cheated, you just did the sums wrong
Thank you, yeah I messed up the sums there for sure thatâs my fault I apologize. I had gotten a bit frustrated with the way the usage is working I think I completely overlooked that part. Greatly appreciate your kindness in pointing that out, others may of bashed me for my ignorant calculation error
The issue I am truly having with the change in structure or however its being calculated is the ârolling 30 daysâ. Since I paid yearly my âbilling periodâ is on the 17th and should reset on that day but isnât. As per a support ticket discussion my count for ârequestsâ is off but affecting how my usage works it seems as the requests are not getting reset. This is what kicked off my initial investigation as I was over 4000 requests used by the 22nd of May and there was no way that was possible.
Since my period doesnât reset I am in usage based pricing because previous month and current month periods are being counted as same rolling month.
I am curious if anyone else is having the similar issue with their Cursor Pro paid for in a yearly rather then monthly like I am?
currently burning 50 an hour on Claude 4 Opus. I donât care, Iâve budgeted the resources all the same, but given we have all these fun unseen systems that may or may not 10% my token usage, and thereâs no predictable or attempt at predicting usage given context in memory, this seems wrong and broken. It certainly doesnât belong where a free mode used to be.
