
Cognition, the company behind Devin, just closed a Series E that values the AI software-engineering startup at $48 billion. The round, announced on September 8, 2026 (UTC), raised over $2 billion and was led by Andreessen Horowitz and Accel, with Founders Fund, General Catalyst, and Avenir joining as existing investors. NVIDIA also participated.
Four months ago, Cognition was worth $26 billion. It has nearly doubled in value since May 2026, when it raised $1 billion in Series D funding. The jump is not hype — it is backed by revenue.
The numbers behind the jump
Cognition's run-rate revenue has grown from $492 million at its Series D in May 2026 to almost $900 million at the Series E announcement. That is an 83% increase in roughly four months, driven by enterprise adoption of Devin, its autonomous software-engineering agent.
| Metric | Series D (May 2026) | Series E (September 2026) |
|---|---|---|
| Round size | $1 billion+ | $2 billion+ |
| Valuation | $26 billion | $48 billion |
| Run-rate revenue | $492 million | ~$900 million |
| Lead investors | Lux Capital, General Catalyst, 8VC | a16z, Accel |
Source: Cognition official announcements, Unite.AI, Bloomberg
Devin now powers engineering teams at NVIDIA in chip design, GE Aerospace in aviation, Citi in financial services, Mercedes-Benz in automotive, and Modal in AI infrastructure. The company says 89% of code committed by its own engineers is committed by Devin, with the rest handled by local agents in Devin Desktop.
What Devin actually does now
The product has expanded beyond "AI software engineer" into a broader agent platform. Devin Auto-Triage takes a first pass at incident investigation. Devin Security Swarm finds and triages vulnerabilities. Devin Automations lets teams configure work to begin from events in Slack, GitHub, Linear, and other systems without opening a chat for each task.
Cognition calls this "the dawn of the self-driving software era." In its vision, agents become proactive by default, software improves itself, and compute budgets self-allocate toward the highest-impact use cases — while human engineers increasingly act as architects who set goals and priorities.
Independent lab, global footprint
Cognition has positioned itself as an independent agent lab, meaning it can choose and combine the models best suited to each job — including its own SWE-1.6 model, which runs at up to 950 tokens per second in Devin Desktop — rather than tying customers to a single provider.
Over the past year, the company opened offices in Washington, D.C., Tokyo, Singapore, London, São Paulo, and Madrid, adding to existing hubs in San Francisco, New York, and Austin.
Why this matters
A $48 billion valuation for a two-year-old startup that sells AI coding agents is a strong signal that the agentic software category is moving from pilot to production. The revenue trajectory — nearly $900 million annualized — suggests enterprises are paying real money for autonomous engineering agents, not just experimenting with them.
For developers, the implication is less about replacement and more about redefinition. Cognition's own numbers show Devin handling the bulk of code commits while humans set direction. That is the model the company is selling, and investors are paying a premium for it.
No comments yet