Cognition raises $2B at $48B valuation for Devin AI

Cognition, the creator of the Devin AI coding assistant, has secured $2 billion in a new funding round that values the startup at $48 billion. Led by prominent venture capital firms Andreessen Horowitz, Accel, Founders Fund, General Catalyst, and Avenir, this massive injection of capital comes just four months after the company was valued at $26 billion. The valuation spike is backed by rapid financial growth, with Cognition reporting that its annualized run-rate revenue has jumped from $492 million in May to $900 million.

This aggressive funding highlights intense investor interest in the AI-assisted programming sector, showing that the market is not a winner-take-all space. Cognition is currently commanding a higher revenue multiple than its rival Cursor did earlier this year. Cursor was in discussions to raise funds at a $50 billion valuation with over $2 billion in annualized revenue before ultimately selling to SpaceX for $60 billion in April. While Cursor is on track to exceed $6 billion in revenue by the end of this year, Cognition is projected to reach between $4 billion and $5 billion in annualized revenue by the end of 2026.

To sustain this growth, Cognition is investing heavily in infrastructure. The startup leases an Nvidia server cluster that costs hundreds of millions of dollars each year, which could drive its total cash burn to $800 million this year. To mitigate these high operational costs, Cognition is training its own proprietary models using open-source alternatives. This strategy aims to reduce its reliance on expensive third-party models from OpenAI and Anthropic, helping the company move closer to breaking even.

For software engineers and enterprise clients, this massive capitalization ensures that Devin will remain a highly supported tool with substantial compute resources. Founded in 2024 by math prodigy Scott Wu, Cognition already serves major enterprise clients including Mercedes-Benz, NASA, Goldman Sachs, and Citi. As Cognition transitions to its own underlying models, developers may see faster execution times and more tailored coding capabilities, while the broader industry benefits from continued competition that prevents a single provider from monopolizing AI-assisted development.