米湯
世界真小. It's a small world!
"
The valuation trajectory on Cursor should break your brain.
$400M in August 2024. $2.5B in December 2024. $9.9B in June 2025. $29.3B in November 2025. Now in discussions at $50B.
That’s 125x in 19 months. For four MIT grads who forked VS Code.
The revenue backs it up. $1B ARR in November. Crossed $2B by February, doubling in 90 days. If that doubling rate holds, they’re approaching $2.5B right now. At $50B, investors are paying somewhere around 20-25x revenue for a company whose growth curve looks more like a crypto chart than a SaaS company.
Here’s what makes the timing so strange. The $50B discussions are happening the same week developers flooded X with posts about switching to Claude Code and citing price as the reason. Cursor’s own data shows 60% of revenue now comes from enterprise, not the individual developers who made it famous.
That split explains the entire fundraise. A developer switches coding tools in an afternoon. An enterprise procurement cycle takes 9 months. Cursor is raising at peak enterprise momentum, before the developer sentiment shift shows up in the renewal data.
And every model provider Cursor depends on is now a direct competitor. Claude Code just crossed $2.5B in run-rate revenue. Google paid $2.4B for Windsurf’s IP and poached its leadership after OpenAI’s $3B acquisition fell apart. Replit raised $400M at a $9B valuation today. The entire AI coding market is being flooded with capital simultaneously.
Cursor’s response: Composer, their in-house model trained on a billion lines of code flowing through the editor daily. The bet is that proprietary usage data creates a moat even if every API provider ships a native IDE.
The window where an independent AI coding platform can exist between the model providers closing in from every side is measured in quarters, not years. $50B is the price of that bet."
--- Aakash Gupta
