Cognition raises $1 billion at $26 billion valuation, CEO touts

Cognition, the San Francisco‑based AI startup behind the autonomous coding agent Devin, announced a $1 billion financing round that lifts its post‑money valuation to $26 billion [1]. The round was led by Lux Capital, General Catalyst and 8VC, with participation from Ribbit Capital, Atreides Management and Founders Fund [2].

Key takeaways

• Cognition raised $1 billion, bringing total capital raised to over $2.5 billion [3].
• The company reports run‑rate revenue of $492 million, a 13‑fold increase from a year earlier [1].
• Enterprise usage of Devin has grown more than tenfold since January 2026, with customers such as Goldman Sachs, Mercedes‑Benz and U.S. defense agencies [3].
• Cognition’s vision is fully autonomous software development, with Devin responsible for 89% of internal code contributions [1].

Funding round and growth metrics

The May 2026 financing round, announced on May 27‑28, was co‑led by Lux Capital, General Catalyst and 8VC, and included new strategic investors Ribbit Capital, Atreides Management and Founders Fund [1][2]. Cognition says the capital will help it stay independent as it scales operations [2]. The company now claims a valuation of $26 billion, more than double its September 2025 valuation of $10.2 billion [3]. With the new money, total funding exceeds $2.5 billion [3].

Cognition reports that its annualized revenue run‑rate has risen to $492 million, up from $37 million a year earlier, representing a 13‑fold jump in twelve months [3]. Enterprise adoption of Devin has increased over tenfold year‑to‑date, with roughly 50% month‑over‑month growth sustained for six months [3]. Major customers span finance, automotive, technology and government sectors, including Goldman Sachs, Citi, Mercedes‑Benz, Dell, Santander, Palantir, NASA and U.S. Army and Navy units [1][3].

AI‑first agent strategy and product roadmap

Cognition positions Devin as an “agent‑first” solution that takes a complete software task—such as a Jira ticket or natural‑language instruction—and executes it in an isolated Linux environment, returning a pull request for human review [3]. The company argues that this autonomous approach differs from IDE‑assistant tools that suggest code inline, and it claims that the agent now contributes 89% of internal code, up from 13% in December 2025 [1][3].

Beyond Devin, Cognition has launched its own model, SWE‑1.6, optimized for speed and efficiency and integrated into tools like Windsurf [1]. The firm’s long‑term vision is fully autonomous software development, where AI agents handle most engineering tasks while human engineers intervene only at key checkpoints [1].

Why it matters

The financing underscores investor confidence in the “agent‑first” paradigm for software development, suggesting a belief that autonomous AI coding agents have a larger addressable market than traditional IDE‑assistant tools. Cognition’s rapid revenue growth and expanding enterprise customer base indicate that large organizations are already experimenting with, and in some cases relying on, autonomous agents to accelerate development cycles and automate security remediation [3]. The next steps will likely involve scaling Devin’s capabilities, expanding model offerings, and navigating the operational challenges of deploying autonomous agents at enterprise scale.