That's quite a war chest to raise in such a short period of time. The market has concentrated with the couple of recent acquisitions. Sounds like likely uses of that cash are
- Buy a few million more users with more generous free tier, as models get cheaper the cost to acquire the marginal customer goes down over time anyway
- Build your own foundation model for coding. tbh I'm skeptical that a company can do this better than the Big 3 AI cos.
- Go to war over enterprise. Do a deal with Deloitte/Accenture and get every single one of their consultants spending 8hrs a day in Cursor. Another flavor: compete head-on with Accenture by making your own service firm that undercuts them and delivers ahead of schedule for once.
luckz 11 hours ago [-]
> Build your own foundation model for coding. tbh I'm skeptical that a company can do this better than the Big 3 AI cos.
They already have their own "tab" model which might not be a very large one but definitely better than most open weight models on short code snippet completion. And for larger agentic LLMs, they can totally start from a pre-trained base model (e.g. deepseek R1) and only do post-train/RL/finetune, which is doable with a small team and their cash reserve.
It's not hard to imagine a good base model (probably deepseek V4?) + cursor's user data leads to a model that surpass sonnet/gemini on coding tasks only.
blitzar 9 hours ago [-]
> Deloitte/Accenture and get every single one of their consultants spending 8hrs a day in Cursor
Unless you change Cursor into a busy work assistant it is functionally useless for consultants.
spacecadet 5 hours ago [-]
Work in consulting. The entire industry has flipped on its head as rates(interest and consulting) have collapsed. Most of the non-techie and somewhat-but-dangerous people(those of us left...) are now spending most of their time vibe coding demos and early prototypes.
dhdodnzjsjs 3 hours ago [-]
> Work in consulting
What is it with western white collar (tech) workers and not using first person pronouns? This sentence took me a couple tries before I realized the “I” was missing.
I see it a lot on this site as well as my own company.
Spivak 1 hours ago [-]
I don't think it's a tech thing, this is a pretty common form in english like "Went to the store. Got some butter."
I think it might feel more common in tech because it's more common in digital communications.
ignoramous 21 hours ago [-]
> Go to war over enterprise ... consultants spending 8hrs a day in Cursor.
And students. Sun's Java push, especially its proliferation as "object-oriented programming language" in CS courses world-wide, might offer a lesson or two.
aduffy 19 hours ago [-]
Students aren’t a market though. Sun and Java was really just marketing to future users
SmellTheGlove 16 hours ago [-]
Isn’t marketing basically spending money now that will result in revenue later?
raspasov 18 hours ago [-]
Future market?
aduffy 16 hours ago [-]
Java only became Java because of enterprise. Getting students to use it was just an exercise in soft power.
solarkraft 14 hours ago [-]
Why do you think enterprise loves it so much?
VectorLock 4 hours ago [-]
Inertia.
bix6 23 hours ago [-]
Wow only a 21.6x rev multiple. The others a month ago were 75x for windsurf acquisition by OpenAI and a 45x on cursors $200M raise at $9B val.
Separately, has anyone gotten through to cursor support? They sent me a welcome email asking for feedback but when I responded nobody answered back.
Edit: added old financing info.
cauliflower2718 20 hours ago [-]
I've gotten an AI generated email back when I asked for support, which did seem to understand my request at a surface level, but was not actually helpful. That's the extent of support I've received from cursor.
hamburglar 14 hours ago [-]
I assume this is our future for almost all support for every product and service.
bionsystem 10 hours ago [-]
Whether it's this or call centers, at the end of the day...
aduffy 21 hours ago [-]
And let's not forget PLTR still hovering just over 100x revenue multiple.
gsibble 23 hours ago [-]
They probably aren't wanting for cash so I wonder why they did this. Maybe a lot of it was secondary sales?
electroly 22 hours ago [-]
On the contrary, they are lighting piles of cash on fire. Their unlimited slow pool is bleeding them dry.
hamburglar 21 hours ago [-]
Re: support, no luck for me either. I even responded to a personal message I got from someone about a glitch with my signup and never heard back. I have no idea if I’m actually paying them.
bix6 21 hours ago [-]
“Please don't hesitate to reach out if there's anything we could do to improve Cursor for you. Making Cursor phenomenal for our pro users is our number one priority.
Best,
The Cursor Devs
”
:(
a13n 23 hours ago [-]
I predict that in 6-12 months we'll all be back on VS Code. I would hate to have Microsoft as a direct competitor, especially in a space they care so much about (developers + AI).
autoconfig 22 hours ago [-]
This thesis has existed since Cursor first started, and the gap between them and VSCode has only widened since then. It’s worth spending some time thinking about why that may be before having such strong conviction about their demise.
rched 21 hours ago [-]
> the gap between them and VSCode has only widened since then
What is in this gap? Do you know of any good resources that outline the features that Cursor provides over VSCode with Copilot?
yunwal 20 hours ago [-]
You can't really name a list of features that cursor has that copilot doesn't. It's more like: Cursor appears to heavily dogfood their features, VSCode's copilot seems to check the feature boxes, but each one sucks to use. The autocomplete popups are jarring. The copilot agent doesn't seem to gather the correct context. They still haven't figured out tool calling. It's really something you have to try rather than look at a checklist of features.
tough 20 hours ago [-]
I think your knowledge is a bit outdated? Cursor definetley still has an edge, but VSCode Github Copilot UI has come a long way and using the same underlying models for both the results are fairly similar and change only in ux niceties
stuff like background agents cursor is way ahead.
Zed Editor is a nice contender too
yunwal 3 hours ago [-]
I tried copilot agent like 3 weeks ago. If that much has changed since then, props to Microsoft.
Zed is very nice, it’s just a totally different workflow. I think people who work in a domain where AI is not particularly strong would be better off with Zed, since Cursor’s way of reviewing edits is a little clumsy.
9rx 19 hours ago [-]
What about on the speed front? VS Code's biggest problem is with how slow it is. I'd already be done and on to the next (and maybe the next thing after that) by the time it finally gets around to things. I like the concept, but I only have so much time in a day.
runako 18 hours ago [-]
If you find VS Code to be slow, you might give Zed a try. I have been using Zed with my Claude API key and it's really something.
keeganpoppen 21 hours ago [-]
have you tried using either of them?
subarctic 21 hours ago [-]
You mean the gap in vscode compatibility?
roxolotl 21 hours ago [-]
What are your thoughts on why it might be?
mritchie712 21 hours ago [-]
* a small, focused team moves faster
* cursor has great taste and that's hard to replicate at MS scale
* Microsoft had allegiance to OpenAI early on which reduced their experimentation with other models
electroly 23 hours ago [-]
Their main supplier (Anthropic) is also a direct competitor (Claude Code). I love Cursor but boy, what a tough place to be in. It's hard to see how it works out for them in the end.
fizx 21 hours ago [-]
Cursor+Gemini MAX is pretty good these days. It seems like Claude Code and C+GM leapfrog each other every month or two.
Cursor has a lot of potential leverage owning the developer and the training data streams and commoditizing the underlying model.
throwaway314155 22 hours ago [-]
Code isn't as strictly competitive as the IDE's are. Code even has solid VS Code integration. It's effectively a plugin, just one that isn't tied to any one IDE.
nikcub 17 hours ago [-]
The new memory feature on Cursor is going to keep me locked in for the foreseeable future. It's _really_ good.
gsibble 23 hours ago [-]
Agreed. There's no way Cursor can stay ahead when they really don't have much of a moat.
Don't get me wrong, I love Cursor but is seems Microsoft could just rip it all off and put it in base VS Code.
outside1234 21 hours ago [-]
Which they have basically done and are closing in on them fast
yunwal 20 hours ago [-]
You and I have not tried the same vscode
ta988 18 hours ago [-]
Have you used copilot recently? It is absolutely useless these days.
bahmboo 15 hours ago [-]
I use Roo Code. I love it. Are Cursor or Copilot or any of the other "front ends" so much better? I guess it's up to me to find out but wonder what others have found. [edit: grammar]
xiphias2 18 hours ago [-]
Since Codex web ui came out I stopped using Cursor and just direct pull requests on Codex web interface. I love it so much, I belive most people will move to this kind of development as models are getting stronger, and the whole agent+user workflow will switch to pull request based development.
It’s not like I’m not using Cursor at all, it just became the 10-20% of my workflow compared to almost 100% before.
archeantus 4 hours ago [-]
Can you elaborate more on this? You’re using a web ide? And you love it?
xiphias2 3 hours ago [-]
Sure, chatgpt.com/codex
habosa 15 hours ago [-]
Ok so OpenAI owns Windsurf now, so I would expect them to cut Cursor off from the really good stuff at some point. Anthropic has Claude Code, they could do the same. Google is a little farther behind but they do have an AI Studio thing that could be viewed as competitive.
Feels like Cursor has to make their own models to guarantee long-term survival? Especially if they’re not going for an acquisition (reports are they turned down OpenAI). Can they make a model that’s good enough for a world where OpenAI / Anthropic / Google all cut them off?
justmarc 10 hours ago [-]
I have a feeling we're getting into an "AI squeeze" soon. So many companies in this sphere are going to have to fight for their lives and or implode.
mythz 13 hours ago [-]
Good to see with that war chest they'll be competing in this space for some time yet - which we'll need more of to keep the pricing down (aka subsidized).
Never tried Cursor since I'm not prepared to leave my IDEs, but still got a full AI toolbox with augmentcode.com, GitHub Copilot, Claude Code and Gemini Code Assist enabled in both my primary VS Code and Rider IDEs.
22 hours ago [-]
adwn 23 hours ago [-]
$900 million?! That's an absolutely insane amount of money. How much of this will go into building the product itself, and how much into converting cash to marketshare? And they raised $100 million not even 5 months ago. Have they already burned through that pile of money?
I mean, their product is good – I'm using and paying for Cursor – but not fantastic. And there's a lot of competition. And the switching cost is relatively low.
gsibble 23 hours ago [-]
I'm wondering if some was used to purchase shares of employees and founders.
bananapub 23 hours ago [-]
they are approximately a service for transferring revenue from paying users to Microsoft or Anthropic, with a small software development project alongside - why would that seem like an insane amount of money?
andrethegiant 19 hours ago [-]
Too busy rolling in money to actually write a substantial announcement?
naiv 23 hours ago [-]
so enough for them to build their own model
stonecharioteer 5 hours ago [-]
I don't understand funding whatsoever. How does a company go from Series B to C in 5 months?
gsibble 23 hours ago [-]
Impressive. Didn't they just raise a bunch recently?
Wow. That's an insane difference in only a few months.
drx 23 hours ago [-]
I thought it was a pretty interesting choice to post the funding announcement as the team and not as the CEO, more companies should probably do that.
moralestapia 15 hours ago [-]
2 million paying users?
Idk, hard to believe.
Wow.
pwm 6 hours ago [-]
I'm assuming a lot is coming from companies where there's fomo that if you don't use the latest and greatest tooling, your competitors will and you'll be left behind.
iamsaitam 21 hours ago [-]
It's very weird that they use their blog to post about this instead of all the feature they've been adding.
Rendered at 17:16:14 GMT+0000 (Coordinated Universal Time) with Vercel.
- Buy a few million more users with more generous free tier, as models get cheaper the cost to acquire the marginal customer goes down over time anyway
- Build your own foundation model for coding. tbh I'm skeptical that a company can do this better than the Big 3 AI cos.
- Go to war over enterprise. Do a deal with Deloitte/Accenture and get every single one of their consultants spending 8hrs a day in Cursor. Another flavor: compete head-on with Accenture by making your own service firm that undercuts them and delivers ahead of schedule for once.
They already have their own "tab" model which might not be a very large one but definitely better than most open weight models on short code snippet completion. And for larger agentic LLMs, they can totally start from a pre-trained base model (e.g. deepseek R1) and only do post-train/RL/finetune, which is doable with a small team and their cash reserve. It's not hard to imagine a good base model (probably deepseek V4?) + cursor's user data leads to a model that surpass sonnet/gemini on coding tasks only.
Unless you change Cursor into a busy work assistant it is functionally useless for consultants.
What is it with western white collar (tech) workers and not using first person pronouns? This sentence took me a couple tries before I realized the “I” was missing.
I see it a lot on this site as well as my own company.
I think it might feel more common in tech because it's more common in digital communications.
And students. Sun's Java push, especially its proliferation as "object-oriented programming language" in CS courses world-wide, might offer a lesson or two.
Separately, has anyone gotten through to cursor support? They sent me a welcome email asking for feedback but when I responded nobody answered back.
Edit: added old financing info.
Best, The Cursor Devs ”
:(
What is in this gap? Do you know of any good resources that outline the features that Cursor provides over VSCode with Copilot?
stuff like background agents cursor is way ahead.
Zed Editor is a nice contender too
Zed is very nice, it’s just a totally different workflow. I think people who work in a domain where AI is not particularly strong would be better off with Zed, since Cursor’s way of reviewing edits is a little clumsy.
* cursor has great taste and that's hard to replicate at MS scale
* Microsoft had allegiance to OpenAI early on which reduced their experimentation with other models
Cursor has a lot of potential leverage owning the developer and the training data streams and commoditizing the underlying model.
Don't get me wrong, I love Cursor but is seems Microsoft could just rip it all off and put it in base VS Code.
It’s not like I’m not using Cursor at all, it just became the 10-20% of my workflow compared to almost 100% before.
Feels like Cursor has to make their own models to guarantee long-term survival? Especially if they’re not going for an acquisition (reports are they turned down OpenAI). Can they make a model that’s good enough for a world where OpenAI / Anthropic / Google all cut them off?
Never tried Cursor since I'm not prepared to leave my IDEs, but still got a full AI toolbox with augmentcode.com, GitHub Copilot, Claude Code and Gemini Code Assist enabled in both my primary VS Code and Rider IDEs.
I mean, their product is good – I'm using and paying for Cursor – but not fantastic. And there's a lot of competition. And the switching cost is relatively low.
Idk, hard to believe.
Wow.