Why the CEO of Cursor is confident that competition from OpenAI and Anthropic won’t overwhelm his startup

Anysphere Prioritizes Innovation Over IPO Plans

Michael Truell, co-founder and CEO of Anysphere—the company behind the popular AI coding assistant Cursor—shared at Fortune’s AI Brainstorm conference that the company has no immediate intentions to go public. Instead, Anysphere is channeling its efforts into expanding Cursor’s capabilities.

Despite achieving $1 billion in annualized revenue as of November and recently securing $2.3 billion in funding at a $29.3 billion valuation, Truell emphasized that product development remains the company’s main focus.

Advancing Proprietary AI Models

Truell highlighted that Cursor’s internally developed large language models (LLMs) are tailored for specific use cases. In a November blog post, Cursor revealed that its proprietary models now generate more code than nearly any other LLM globally.

During the event, Truell addressed questions about competing with major LLM providers like OpenAI and Anthropic, who also offer AI coding solutions. He compared their offerings to “concept cars,” while positioning Cursor as a fully realized production vehicle. “Rather than just assembling an engine and a prototype, we integrate the best intelligence from various providers and enhance it with our own specialized models, resulting in a comprehensive tool and user experience,” Truell explained.

Strategic Shifts and Industry Challenges

Cursor’s reliance on both its own and competitors’ models has sparked speculation among Silicon Valley investors, especially after reports surfaced that OpenAI considered acquiring Anysphere earlier this year—a proposal Anysphere declined.

Investors noted that AI coding platforms were struggling with profitability due to high costs from model providers. In response, Cursor transitioned to a usage-based pricing model in July, directly passing API fees to users. This shift away from an all-inclusive subscription led to some customer dissatisfaction, as unexpected charges emerged.

Addressing the pricing controversy, Truell explained, “Cursor has evolved from answering quick coding questions to handling hours of work. This required us, and others in the industry, to adopt a consumption-based pricing model.”

Enterprise Solutions and Future Directions

Truell shared that Anysphere is developing tools similar to cloud cost-management systems, enabling organizations to track and control their engineering expenses. “We have a dedicated team building features like spend controls, billing groups, and enhanced visibility for enterprises,” he said.

Looking ahead, Cursor is concentrating on two key areas for the coming year:

• Advanced Agentic Capabilities: The goal is for Cursor to autonomously complete complex, end-to-end tasks that are simple to describe but challenging to execute—such as intricate bug fixes that might otherwise take weeks of manual effort.
• Team-Oriented Solutions: Cursor aims to shift its focus from individual developers to supporting entire teams, reflecting its growing enterprise customer base.

Beyond cost management, Truell wants Cursor to assist with more stages of the software development process, not just code writing. For example, Cursor’s code review tool is already being used to analyze every pull request, regardless of whether it was written by a human or AI.

“You’ll see us rolling out more features designed to help teams as a whole,” Truell promised.

Competitive Landscape and Industry Collaboration

Meanwhile, major tech companies are also preparing for a future dominated by sophisticated AI agents. Amazon, for instance, recently introduced a coding tool capable of running for extended periods.

This week, leading AI organizations—including Anthropic, OpenAI, Microsoft, and AWS—formed a new consortium under the Linux Foundation to establish open-source standards for agentic interoperability. Notably, they contributed significant projects such as Anthropic’s Model Context Protocol (MCP).

While Anysphere’s roadmap may not immediately propel it ahead of its largest competitors, these initiatives are expected to keep the company competitive in the rapidly evolving AI coding space.