Pomegra
Cognition AI closed a $2 billion Series E at $48 billion, led by Andreessen Horowitz and Accel, as Devin's annualized revenue closes in on $1 billion.
Key Takeaways
- Cognition raised $2B+ in a Series E co-led by Andreessen Horowitz and Accel at a $48B post-money valuation, announced September 8, 2026.
- The valuation nearly doubled from $26B in May 2026, when Cognition closed a $1B Series D just four months earlier.
- Devin's annualized revenue grew from $492M at the Series D to roughly $900M by September 2026.
The $48 Billion Number
San Francisco-based Cognition AI raised more than $2 billion on September 8, 2026, in a Series E that puts the company's post-money valuation at $48 billion. Andreessen Horowitz and Accel co-led as new investors. Founders Fund, General Catalyst, and Avenir returned alongside a syndicate that includes Benchmark, Bessemer Venture Partners, Kleiner Perkins, Greylock, Lightspeed, NVIDIA, and more than two dozen other firms. Cognition's Devin AI coding platform now generates roughly $900 million in annualized revenue, up from $492 million when the company raised at $26 billion in May.
That four-month gap between major rounds is the first thing worth examining. Cognition didn't need to raise this quickly. Doing so at nearly double the prior valuation tells you something about both demand and trajectory - though not necessarily everything the press release implies.
What Does Devin Actually Do?
Devin is an AI software engineer. It takes on coding tasks autonomously - bug fixes, repository migrations, test pipelines - inside its own sandboxed development environment. It reads codebases, writes and revises code, executes tests, and returns pull requests for human review. The pitch is simple: treat it like a junior engineer who never needs onboarding and works continuously.
Most AI coding tools operate as function-level autocomplete. Devin's design targets entire repositories and multi-step workflows. That distinction is the basis for Cognition's claim that its addressable market isn't just developer tooling - it's software engineering labor at scale. Founded in November 2023 by CEO Scott Wu, CTO Steven Hao, and CPO Walden Yan, the company went from zero revenue to near-$1 billion annualized in under three years, a pace that few enterprise software businesses have matched.
Whether the $900 million run rate is durable is another question. Customer concentration, renewal rates, and billing structure remain undisclosed. The number is real; what it implies about retention and expansion is not yet visible from the outside.
What Does This Round Imply About the Last One?
A $48 billion valuation on roughly $900 million in annualized revenue implies a revenue multiple of about 53x. That is aggressive even by current AI sector standards, where 30-40x has become normalized for high-growth platforms. The May Series D, at $26 billion on $492 million ARR, implied a multiple in the same range. So the market is not re-rating Cognition's growth story - it is simply applying a consistent premium to a larger revenue base.
The 85% ARR growth between rounds gives investors a defensible rationale. At that pace, Cognition crosses $1 billion in annualized revenue before year-end. For a company not yet three years old, that trajectory justifies the multiple more convincingly than market-size narratives alone.
The breadth of the investor syndicate is its own signal. Rounds this oversubscribed, with names this recognizable, tend to reflect allocation competition as much as conviction. When Benchmark, Kleiner Perkins, and a16z are all in the same cap table, you are looking at a statement about market position as much as a financial bet.
Competitive Context
The AI coding market is now crowded in ways it wasn't when Cognition launched. GitHub Copilot remains the incumbent with the largest installed base. Cursor has built a strong following among individual developers. Newer entrants continue to emerge as underlying model capabilities improve.
Cognition's differentiation rests on Devin's ability to handle longer, more complex tasks than point-and-click coding assistants. The enterprise version of that story - where companies want a system that can autonomously handle full engineering workflows, not just suggest completions - is where Cognition is making its commercial case. The $900 million ARR number suggests the argument is landing, at least for now.
Microsoft via GitHub and Google are the structural competitors that matter most long-term, given their distribution advantages. Neither has matched Devin's autonomous agent framing yet, but both have the resources to try.
Outlook
Cognition has capital, a verified revenue trajectory, and the most densely syndicated investor base in its history. The path to $1 billion in annualized revenue is arithmetically close. The harder question is whether a 53x multiple survives any deceleration in that growth rate - and whether the current revenue pace reflects genuine enterprise adoption or the early-adopter surge common to new AI categories. The next twelve months will answer that question more clearly than any funding announcement can.
