# Why is a simple edit eating 100,000+ tokens? Let’s talk about this **URL:** \ **Category:** Discussions\ **Created:** [July 19, 2025, 9:08am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025 "2025-07-19T09:08:06Z")\ **Posts on this page:** 20\ **Page:** 1
**Author:** ![Mohamed\_Khafagy](https://avatars.discourse-cdn.com/v4/letter/m/c67d28/32.png) [@Mohamed\_Khafagy](https://forum.cursor.com/u/Mohamed_Khafagy)\ **Post date:** [July 19, 2025, 9:08am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/1 "2025-07-19T09:08:06Z")
Hey all — Really enjoying Cursor Pro and loving how it helps with AI-powered editing. But… something’s not adding up when it comes to **token usage** and I think it’s worth discussing openly here. Here’s what happened: 👉 I edited a single file that was about **7,500 tokens in total size** (see screenshot below). 👉 After that one edit, Cursor’s dashboard shows that I used **over 107,000 tokens for that one transaction**! ![cursor2](https://us1.discourse-cdn.com/cursor1/original/3X/2/2/22d3762f6277931c3a8d710a5642ab2bc950fa59.png) ![cursor1](https://us1.discourse-cdn.com/cursor1/original/3X/9/b/9b5bf1268671f154e3111f7bdc9572d7fae67d18.png) That’s not a typo — one simple edit, on one moderate-sized file, burned through (107k tokens in one go). This honestly shocked me. When I was thinking about building something like this myself, I would have expected this workflow to consume that much over the (course of an entire week) — not a single edit. And this wasn’t some crazy large context or complex prompt — just normal editing. Even accounting for caching, summarization, etc., I can’t help but feel like (there’s a lot of invisible overhead going on here). This matters because: - If every small edit costs this much in tokens, it could explain why people hit limits faster than expected. - Users might not realize that a “minor” edit can result in massive token charges behind the scenes. - I’d bet there’s at least 70%+ potential savings possible here without impacting quality — just with more efficient prompt design and context management. I’d be interested to hear if others have observed similar patterns, and would appreciate any insight from the Cursor team on how this token usage is calculated for these types of edit actions. My goal in raising this is simply to understand how the system works and whether there are future plans for further optimization. Thanks again to the Cursor team for the great product — this just seemed like an important topic to discuss constructively. ---
**Author:** ![Artemonim](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/artemonim/32/56113_2.png) [@Artemonim](https://forum.cursor.com/u/Artemonim)\ **Post date:** [July 19, 2025, 9:34am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/2 "2025-07-19T09:34:32Z")
> [@Mohamed\_Khafagy](#): > > see screenshot below This tokenizer can be wrong by at least 10% in relation to different LLMs. > [@Mohamed\_Khafagy](#): > > After that one edit - One `edit` or one `request`? - Is the entire workspace a single file? ---
**Author:** ![Mohamed\_Khafagy](https://avatars.discourse-cdn.com/v4/letter/m/c67d28/32.png) [@Mohamed\_Khafagy](https://forum.cursor.com/u/Mohamed_Khafagy)\ **Post date:** [July 19, 2025, 10:23am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/3 "2025-07-19T10:23:16Z")
Yes, but this was around 10x the tokens, not just a 10% variance from tokenizer differences, so I don’t think it’s a tokenizer mismatch issue. This was a single API request using “Claude 4” (not Fast Edit), on a single file that totaled approximately 7,500 tokens. However, Cursor’s dashboard shows that this request consumed over 107,000 tokens — suggesting that the entire file content may have been sent repeatedly within that one request. From what I can tell, this implies there might not have been effective caching or summarization applied at the prompt layer, at least for this workflow. It feels like the file could have been included multiple times in prompt chains or intermediate calls during processing, which adds significant hidden token overhead. ---
**Author:** ![Artemonim](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/artemonim/32/56113_2.png) [@Artemonim](https://forum.cursor.com/u/Artemonim)\ **Post date:** [July 19, 2025, 10:30am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/4 "2025-07-19T10:30:13Z")
> [@Mohamed\_Khafagy](#): > > This was a single API request using “Claude 4” (not Fast Edit), on a single file that totaled approximately 7,500 tokens. How did you know there was only one “API request”? Was it executed in Manual mode? Did the agent use the tools only once or didn’t use them at all? ---
**Author:** ![Mohamed\_Khafagy](https://avatars.discourse-cdn.com/v4/letter/m/c67d28/32.png) [@Mohamed\_Khafagy](https://forum.cursor.com/u/Mohamed_Khafagy)\ **Post date:** [July 19, 2025, 10:35am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/5 "2025-07-19T10:35:04Z")
Sorry for that i mean it was single request using claude 4, not single api request,and also i selected the lines i want to edit and did not leave it to read whole file to find it, and it take 10x tokens than file size ---
**Author:** ![Artemonim](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/artemonim/32/56113_2.png) [@Artemonim](https://forum.cursor.com/u/Artemonim)\ **Post date:** [July 19, 2025, 10:39am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/6 "2025-07-19T10:39:15Z")
> [@Is it possible to export the entire chat?](https://forum.cursor.com/t/is-it-possible-to-export-the-entire-chat/118837/2): > > hi @Artemonim yes in the Chat you have the option to export current chat as MD file. You can also load from history other chats and export. Alternatively you can install SpecStory extension which automatically exports each chat into a folder .specstory also as MD files. Export the chat and paste it here ---
**Author:** ![Mohamed\_Khafagy](https://avatars.discourse-cdn.com/v4/letter/m/c67d28/32.png) [@Mohamed\_Khafagy](https://forum.cursor.com/u/Mohamed_Khafagy)\ **Post date:** [July 19, 2025, 11:10am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/7 "2025-07-19T11:10:15Z")
[cursor\_prevent\_same\_gender\_chat\_interac.md.txt](https://forum.cursor.com/uploads/short-url/hiC1mPJZuHfklt0gOAJxjlDwocY.txt) (5.8 KB) This was less than 50 lines of code, but the output consumed 107K tokens. Honestly, when I was planning an editor for my internal company use, I wasn’t expecting to send the entire code 10 times in a single request, that kind of behavior feels really strange. It seems like there’s a significant opportunity here for Cursor to restructure how prompts and code are sent to expensive models like Claude Sonnet or Opus, possibly from the ground up. If it’s helpful, happy to volunteer some time to think through this with the team ---
**Author:** ![Artemonim](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/artemonim/32/56113_2.png) [@Artemonim](https://forum.cursor.com/u/Artemonim)\ **Post date:** [July 19, 2025, 11:21am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/8 "2025-07-19T11:21:39Z")
Well, yes, 3 edit\_tool shouldn’t have read 8k tokens 14 times… ---
**Author:** ![Mohamed\_Khafagy](https://avatars.discourse-cdn.com/v4/letter/m/c67d28/32.png) [@Mohamed\_Khafagy](https://forum.cursor.com/u/Mohamed_Khafagy)\ **Post date:** [July 19, 2025, 1:12pm UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/9 "2025-07-19T13:12:04Z")
Hope they can give explanation here ---
**Author:** ![SeanPM-21](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/seanpm-21/32/71698_2.png) [@SeanPM-21](https://forum.cursor.com/u/SeanPM-21)\ **Post date:** [July 19, 2025, 1:48pm UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/10 "2025-07-19T13:48:39Z")
A-lot of people are having this same issue. I messaged Cursor and they said that I should switch to Auto for the time being. Like I paid $200 to not have access to the models in my plan lol. If they don’t fix this soon, ill either try to force a refund, or chargeback. ---
**Author:** ![Mohamed\_Khafagy](https://avatars.discourse-cdn.com/v4/letter/m/c67d28/32.png) [@Mohamed\_Khafagy](https://forum.cursor.com/u/Mohamed_Khafagy)\ **Post date:** [July 19, 2025, 4:28pm UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/11 "2025-07-19T16:28:13Z")
Looks very strange really, why do they use this sum of tokens even? I think they have long sequence of calls and everytime the send full code again, they can - with small manipulation - sent only the structure and avoid such big usage, lets wait for the explanation ---
**Author:** ![0xHACKS](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/0xhacks/32/9577_2.png) [@0xHACKS](https://forum.cursor.com/u/0xHACKS)\ **Post date:** [July 20, 2025, 12:00am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/12 "2025-07-20T00:00:43Z")
[MOD - Condor - off topic, please use more relevant threads] ---
**Author:** ![luke\_thomas](https://avatars.discourse-cdn.com/v4/letter/l/b782af/32.png) [@luke\_thomas](https://forum.cursor.com/u/luke_thomas)\ **Post date:** [July 20, 2025, 12:09am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/13 "2025-07-20T00:09:44Z")
Use auto first, because it is free, and then if it doesn’t work revert to the old version and then try a specific model. Even better, automate the process and track performance differences. ---
**Author:** ![kristinaOA](https://avatars.discourse-cdn.com/v4/letter/k/90ced4/32.png) [@kristinaOA](https://forum.cursor.com/u/kristinaOA)\ **Post date:** [July 20, 2025, 12:22am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/14 "2025-07-20T00:22:26Z")
Thanks, but NO! I personally will not use auto until it is more reliable! Only really bad experiences when switching to auto for code editing… ---
**Author:** ![luke\_thomas](https://avatars.discourse-cdn.com/v4/letter/l/b782af/32.png) [@luke\_thomas](https://forum.cursor.com/u/luke_thomas)\ **Post date:** [July 20, 2025, 12:34am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/15 "2025-07-20T00:34:26Z")
It’s a lot better than it used to be. I spent most of the day giving identical prompts to claude and auto, and auto produced working code 15 percent faster on average than claude took to produce very buggy code even when handed the problem. They both started with empty fields and they both were given the exact same prompts. ---
**Author:** ![kristinaOA](https://avatars.discourse-cdn.com/v4/letter/k/90ced4/32.png) [@kristinaOA](https://forum.cursor.com/u/kristinaOA)\ **Post date:** [July 20, 2025, 12:59am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/16 "2025-07-20T00:59:49Z")
I can beat you when it comes to tokens consumed… Agent Claude 4 Sonnet Input 42 Output 2,383 Cache write 63,111 **Cache Read 474,254** **Total Tokens 539,790** luckily was within the plan. But this makes me nervous about pricing and usage cost once I run out of the plan (and this month I almost haven’t touched Cursor) And all that it is achieving,- is for me to look at going Claude path (since that’s pretty much only model I use and seems to work). I tried Grok 4,- froze in the middle every single time! ---
**Author:** ![beru](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/beru/32/71130_2.png) [@beru](https://forum.cursor.com/u/beru)\ **Post date:** [July 20, 2025, 1:09am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/17 "2025-07-20T01:09:29Z")
wait. so is he right that the token number in the dashboard is inflated by 14 times ? ---
**Author:** ![Mohamed\_Khafagy](https://avatars.discourse-cdn.com/v4/letter/m/c67d28/32.png) [@Mohamed\_Khafagy](https://forum.cursor.com/u/Mohamed_Khafagy)\ **Post date:** [July 20, 2025, 1:51am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/18 "2025-07-20T01:51:28Z")
Unfortunately,yes ---
**Author:** ![Mohamed\_Khafagy](https://avatars.discourse-cdn.com/v4/letter/m/c67d28/32.png) [@Mohamed\_Khafagy](https://forum.cursor.com/u/Mohamed_Khafagy)\ **Post date:** [July 20, 2025, 1:58am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/19 "2025-07-20T01:58:42Z")
So it sends your code about 8 times to clause before get you the output, why? and why it sends complete code again and again, its very easy to change their prompt structure and summerize code to avoid this, I think cursor is depending 100% in selected AI when you select model manually, so it sends all your code to selected model again and again instead of passing it first to smaller model to optimize it. ---
**Author:** ![Mohamed\_Khafagy](https://avatars.discourse-cdn.com/v4/letter/m/c67d28/32.png) [@Mohamed\_Khafagy](https://forum.cursor.com/u/Mohamed_Khafagy)\ **Post date:** [July 20, 2025, 2:01am UTC](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025/20 "2025-07-20T02:01:16Z")
For me auto is giving long and sometime strange solutions, it works like junior developer who you should review his code very carefully [Next page](https://forum.cursor.com/t/why-is-a-simple-edit-eating-100-000-tokens-let-s-talk-about-this/120025.md?page=2)