He says in the video with his SkateBench that he doesn’t get why you’d need a expensive microscope if you can just hammer nails with a cheap hammer. But he doesn’t say that you’d take a hammer to hit nails anyway.
He also had a price of $72/day there - I got about $8/h when using Gemini 2.5 Pro (if Gemini is running almost constantly at that hour).
Agreed. It’s a common trap with programmers to prematurely optimize code, especially at the sacrifice of readability, maintainability, and development time. No one really cares how you code something.
There are countries where you can earn (conditionally) $80, $8 and $2 per hour for roughly the same work. And the price difference of (conditionally) +$4 will be different in each of these countries. And if you take a random Cursor user, i think, you’ll be closer to the $8 bar than the $80 bar.
Rich and powerful people are not stupid. By definition in fact. if you can’t understand why they’re doing something, it’s not a sign that they’re going to go bust, it’s a sign that you don’t understand what they’re really doing.
There have never before been intelligent machines, so nobody can predict what happens next, but you can be sure that whoever comes out on top, will be the smartest one(s). You don’t get “smartest” if you can’t learn. You can’t learn if you don’t have customers. When it costs more than you’re paying: you’re the product.
Best explanation yet. Everyone who is like “well I’m going to claude code, just canceled my cursor subscription. Cursor is going to go bankrupt because they hired some idiot to make these pricing policies.” thinks this is some rinky d1nk start up. Everything Cursor is doing is to succeed as a business and they have the money and incentive to not only survive but to dominate a large part of the market.
Cursor is closely linked to big tech players and most likely will become a monster, that I will jump ship from at some point. But they are definitely here to succeed.
new funding to improve Cursor, $900 million at a $9.9 billion valuation
Timeline of Investment of key players (ChatGPT):
# Anysphere (Cursor IDE) — Ownership & Investment Timeline (2022–2025)
## Founders
- **Michael Truell (CEO)**
- **Sualeh Asif**
- **Arvid Lunnemark**
- **Aman Sanger**
---
## Pre-Seed — April 2022 (~$400k)
- Early funding, investors not publicly disclosed.
---
## Seed — October 11, 2023 ($8M)
- **Lead:** OpenAI Startup Fund
- **Notable Angel Investors:**
- Nat Friedman (ex-GitHub CEO)
- Arash Ferdowsi (Dropbox cofounder)
- Other unnamed angel investors
---
## Series A — August 22, 2024 ($60M)
- **Institutions:**
- Andreessen Horowitz (a16z)
- Thrive Capital
- OpenAI (participated)
- **Individual/Strategic Angels:**
- Jeff Dean (Google DeepMind Chief Scientist)
- Noam Brown (Meta AI researcher)
- Founders of:
- Stripe
- GitHub
- Ramp
- Perplexity
- OpenAI
---
## Series B — January 16, 2025 ($105M, ~$2.6B valuation)
- **Institutions:**
- Thrive Capital
- Andreessen Horowitz (a16z)
- Benchmark
- Participation from existing investors (OpenAI, angels, etc.)
---
## Series C — June 6, 2025 ($900M, ~$9.9B valuation)
- **Lead:** Thrive Capital
- **Institutions:**
- Accel
- Andreessen Horowitz (a16z)
- DST Global
- **Existing investors** (OpenAI, Benchmark, angels) continued participation
---
# Key Points
- **Andreessen Horowitz (a16z):** Participated in Series A, B, and C → consistent major investor.
- **Thrive Capital:** Most influential institutional backer, leading Series C.
- **OpenAI Startup Fund:** Early (seed) lead investor; role diluted in later rounds.
- **Benchmark, Accel, DST:** Joined at later stages.
- **High-profile angels:** Nat Friedman, Arash Ferdowsi, Jeff Dean, Noam Brown, plus founders of major tech companies.
There is either a massive bubble going on or all these AI companies are losing money, as @Chris_Drake said, to learn from their customers, and at some point they are going to start cashing in (like what we see with Cursor’s price changes). I think the prices we see now are way too low for the value even as solo devs are getting from it, let alone the value large companies will get when they start replacing entire departments with AI. Thee companies will pay much, much more than what Cursor is currently charging to get that value. And yea, Cursor as we know it may no longer exist, but they will either be consumed by some larger company or transform into something else (which it may already be on its way).
ChatGPT:
🚀 Growth Summary
2022 (Pre-seed): ~<$10M valuation (est.)
2023 (Seed): ~$30–50M valuation (est.)
2024 (Series A): ~$400–600M valuation (est.)
Jan 2025 (Series B): ~$2.6B valuation (disclosed)
Jun 2025 (Series C): ~$9.9B valuation (disclosed)
📊 Overall growth:
From <$10M → ~$9.9B in ~3 years = ~1,000× increase.
That’s an exceptionally steep curve, even by Silicon Valley AI standards.
Cursor is clearly the chosen one.
Of course anthropic developing a cursor like tool would be a competitor:
Exactly, as a solo dev builder myself, its easy to recognize ,if the project is not viable, it’s going to be priced out of cursor, very soon. Anysphere is not to be underestimated.
My cursor account is an ongoing business expense priority.
