The Last Invention is AI

Episode: Cursor Hits Major Milestone With $2.3B Round
Host: Jaden Schaefer
Date: November 15, 2025

Episode Overview

In this episode, host Jaden Schaefer delves into the staggering news that Cursor, an AI-powered coding assistant, has secured $2.3 billion in new funding just five months after its last major round. The discussion unpacks Cursor's meteoric rise, its strategy in a fiercely competitive market, and the technical innovations powering its growth. Jaden also explores how Cursor's move to develop its own foundational AI model could reshape the coding ecosystem and what this means for developers, investors, and the broader AI landscape.

Key Discussion Points and Insights

1. Cursor’s Explosive Funding and Soaring Valuation

• Cursor raises $2.3 billion, more than doubling its previous valuation from $9.9B to $29.3B ([00:30]).
• The last funding round just five months ago was already impressive at $900 million.
• Co-led by Accel (existing investor) and Coatue (new to the cap table), with heavyweights Nvidia and Google joining strategically.
• Nvidia’s participation is strategic—as Jaden notes, “Nvidia is basically handing out money to the top AI use cases as they know the money is going to come straight back to them as more compute will be needed.” ([02:00])

2. Cursor’s New AI Model: Composer

• Michael Truel, Cursor’s CEO, indicates all the new funding will fuel the development of Composer, their in-house AI model, released in October ([03:00]).
• Unlike many competitors reliant on OpenAI or Anthropic for core models (and consequently, at the mercy of their politics and fees), Cursor has moved to full independence by building its own. Jaden highlights this as “obviously why they’re able to raise so much money” and underscores its centrality to Cursor’s runaway success ([04:00]).

3. Competitive Landscape for AI Coding Assistants

• Cursor is the top coding tool in terms of users, but faces intensifying competition from OpenAI and Anthropic’s own coding products (e.g., Claude Code).
• Cursor boasts a massive user base: “over 1 million daily users” and “tens of thousands of enterprises”—including OpenAI, Instacart, and Salesforce ([05:20]).
• Notably, even employees at OpenAI reportedly use Cursor, despite the company making a rival tool ([06:00]).

4. Technical Innovation: Integration and Performance

• Cursor is built atop Microsoft’s open-source VS Code editor and enhanced with its own LLM, Composer ([07:00]).
• Composer is described as a “mixture of experts” system—“when there is a question it sends it to a number of quote unquote experts inside of the AI model that each determine which can answer the question best.” ([07:30])
• Claims a fourfold speed increase over comparable LLMs, with most coding tasks completed in under thirty seconds ([08:30]).
• Real-world impact: Faster responses are a massive productivity boost for developers, especially compared to slow cloud-based code generation (“cloud code … can sit there for, for 10 or 15 minutes working on the code base”) ([09:20]).

5. Deep Technical Choices: No CUDA Libraries

• Jaden explains that most LLMs use CUDA libraries for parallelized code execution on Nvidia GPUs. Cursor, however, “did not use any CUDA libraries while they were building Composer. They said that they implemented the models kernels using pxt. This is kind of a, what’s called a low level machine language.” ([10:30])
• This “in-house” approach reportedly gives Cursor over a 3x performance boost in critical components. This engineering feat further explains Nvidia’s deep interest and investment.

Memorable Quotes & Notable Moments

• On funding and growth:
  “What’s crazy is this coming five months after its previous round, it just continues to, I mean, just absolutely absorb venture capital. Its valuation continues to rise.” — Jaden Schaefer ([00:16])

• On Nvidia’s involvement:
  “Nvidia is basically handing out money to the top AI use cases as they know the money is going to come straight back to them as more compute will be needed.” — Jaden Schaefer ([02:00])

• On Cursor’s strategic independence:
  “So often with these AI companies, they get into this tricky situation where they're reliant on either OpenAI or Anthropic... The fact that they've actually developed their own AI model is obviously why they're able to raise so much money.” — Jaden Schaefer ([04:00])

• On technical performance:
  “This all happens in the background very quickly, so you're not going to notice it. You just ask a question, it will give you a response. But that's how it's working… It can complete a lot of coding tasks in under 30 seconds.” — Jaden Schaefer ([08:00])

• On Cursor’s engineering approach:
  “Cursor says that they did not use any CUDA libraries while they were building Composer and they said that they implemented the models kernels using pxt. … That approach apparently has helped Cursor achieve more than 3x performance increase across a bunch of their components.” — Jaden Schaefer ([10:45])

Timestamps for Key Segments

• [00:00] Cursor’s explosive funding round and new valuation
• [01:50] Strategic investors (Nvidia, Google, Coatue) and the VC landscape
• [03:00] CEO Michael Truel on funding use—Composer AI model
• [04:00] Importance of in-house models and market positioning
• [05:20] Cursor’s user base and competitive context (with OpenAI/Anthropic)
• [07:00] Cursor’s technical stack—VS Code, LLM, and Composer specs
• [08:30] Performance differences vs. competitors—why speed matters
• [10:30] Technical deep dive: building without CUDA, use of “pxt”
• [12:00] The strategic angle for Nvidia and implications for the GPU market

Episode Tone

Jaden’s tone throughout is energetic, impressed, and occasionally incredulous at the pace and scale of Cursor’s progress—reflecting both excitement for technological leaps and an appreciation for the competitive, high-stakes race unfolding in AI. The insights are delivered in a straightforward, approachable manner with technical explanations made accessible—even as deeper geekery about GPUs and model architecture emerges.

For Listeners New to Cursor and this AI Race

• This episode is an accessible but deep dive into how and why Cursor has quickly become a leader in the coding assistant space, the strategic chess moves of tech titans (from Nvidia to Google), and how foundational, in-house AI models may remake the platform landscape for developers and enterprises alike.
• Whether you’re an engineer, investor, or just fascinated by AI, the episode explains both the business rationale and the technical underpinnings steering Cursor’s rise—and why these moves are reverberating across the software world.