Artificial intelligence coding startup Cognition has raised more than $2 billion in fresh funding, pushing its valuation to a staggering $48 billion as investor appetite for AI-powered software development continues to accelerate.
The Series E round was led by new investors Andreessen Horowitz and Accel, alongside existing backers Founders Fund, General Catalyst and Avenir. A long list of other investors, including NVIDIA, Benchmark, Bessemer Venture Partners, Greylock and Lightspeed, also participated.
The deal is particularly striking because Cognition was valued at $26 billion only four months ago.
From $26 Billion to $48 Billion in Four Months
Cognition raised more than $1 billion at a $26 billion valuation in May 2026. Its latest round means investors have now almost doubled the value they place on the company in a matter of months.
Behind that extraordinary rise is equally rapid revenue growth.
Cognition says its annualised run-rate revenue has climbed from $492 million in May to nearly $900 million. That translates to roughly $75 million in monthly revenue at its current run rate.
The numbers provide another indication of just how quickly AI coding tools are moving from experimental products into major commercial businesses.
Meet Devin, Cognition’s ‘AI Software Engineer’
At the centre of Cognition’s growth is Devin, its autonomous AI software-engineering agent.
Rather than simply suggesting individual lines of code, Devin is designed to take on broader engineering tasks, including planning work, writing and testing code, investigating incidents and helping address security vulnerabilities.
The bigger idea is to allow human engineers to focus increasingly on architecture, strategy and complex decision-making while AI agents handle more of the execution.
Cognition was founded in 2024, making its rise to a $48 billion valuation particularly remarkable.
Some of the World’s Biggest Companies Are Already Customers
Cognition says Devin is already being used inside major global companies across several industries.
Its customers include NVIDIA in chip design, GE Aerospace in aviation, Citi in financial services, Mercedes-Benz in automotive and Modal in AI infrastructure.
That enterprise adoption is important because the race to dominate AI coding is increasingly moving beyond individual developers experimenting with assistants.
Large companies are beginning to look at AI agents as a way of increasing the amount of software their existing engineering teams can build and maintain.
Andreessen Horowitz, one of Cognition’s new investors, says Devin now writes more than 90% of Cognition’s own production code, up from 13% a year earlier. The venture capital firm also pointed to deployments at companies including Mercedes-Benz and Rivian as evidence of the technology’s potential.
The AI Coding Race Is Getting Expensive
Cognition isn’t operating in an empty market.
AI-assisted software development has become one of the most competitive areas of the broader artificial intelligence boom, with developers increasingly able to use AI to generate, debug and manage code.
The enormous valuations being attached to companies in the sector suggest investors believe AI coding could become one of the largest commercial applications of generative AI.
But building these systems isn’t cheap.
Cognition reportedly leases an NVIDIA server cluster costing hundreds of millions of dollars annually, while its cash burn could reach around $800 million this year.
That means the company’s near-$900 million revenue run rate is impressive, but so too is the amount of capital required to compete at the frontier of AI.
A $48 Billion Bet on the Future of Software Development
Cognition’s latest funding round ultimately represents more than another huge AI investment.
It is a multibillion-dollar bet on a fundamental change in who — or what — writes the world’s software.
For decades, software development has largely meant humans manually translating ideas and requirements into code. Companies such as Cognition are betting that increasingly autonomous AI agents will take over much of that execution.
The company argues that global demand for software already exceeds the industry’s capacity to build it, leaving products unfinished, upgrades delayed and vulnerabilities unresolved because engineering time is limited.
If AI agents such as Devin can close that gap, the market could be enormous.
Investors have now put a $48 billion price tag on Cognition’s attempt to prove it.
