Cognition’s $40B Talks: Why Investors Are Chasing an AI Coding Agent

From $26B to $40B in Three Months

Cognition, the maker of AI coding agent Devin, is in early talks with investors for a new funding round that could value the company at $40 billion or more, according to PYMNTS’ reporting, which cites Bloomberg. That would represent more than a 50% jump from the $26 billion valuation Cognition secured less than three months earlier, when it raised $1 billion in a round reported by TechCrunch. PYMNTS reports the new round is expected to bring in more than $1 billion again, though the exact terms remain in early negotiation, with The AI Insider confirming the discussions were still at an “early” stage as of its August 13, 2026 reporting.

The revenue growth behind the new valuation target is concrete and specific. TechCrunch reports Cognition’s annualized revenue run rate reached $492 million as of the May 2026 round, and PYMNTS reports the company’s current run rate is now “nearly $1 billion” — approximately double the May figure. That kind of revenue doubling in roughly three months is the central justification investors are using to support a valuation jump of similar magnitude.

Who Is Actually Using Devin

Devin’s customer list has expanded from a purely private-sector story to a mixed public-private one. TechCrunch names Mercedes-Benz, NASA, and Goldman Sachs as enterprise customers as of the May 2026 round, while PYMNTS’ more recent reporting adds Citi, the U.S. Army, and the U.S. Navy to the customer list. The presence of two U.S. military branches alongside NASA signals that Cognition has moved beyond software-company early adopters into government and defense-adjacent procurement — a customer segment that typically involves longer sales cycles and more rigorous security vetting, which makes its appearance in Cognition’s customer base a meaningful validation signal for enterprise AI credibility.

TechCrunch also reports a specific usage metric: enterprises using Devin grew their usage by 50% month-over-month for the six months preceding the May 2026 round, a figure independently corroborated by The AI Insider’s reporting. Cognition co-founder and CEO Scott Wu — described by The AI Insider as a “prominent young programmer” — has been explicit about positioning the product, telling TechCrunch that “Devin is not being sold as a human replacement,” and that the agent instead handles longer-term technical tasks that developers often avoid. The AI Insider’s version of the same positioning frames Devin as designed to “handle repetitive, lower-priority engineering tasks” such as modernizing outdated software or migrating applications between platforms — a framing aimed at defusing the displacement anxiety that has dogged AI coding tools broadly.

The Wider AI Coding Land Grab

Cognition’s fundraising talks are not happening in isolation — they’re part of a broader wave of high-value deals in the AI coding and agent space that PYMNTS’ reporting places in context. The same period saw reports of SpaceX acquiring Cursor in a deal valued at $60 billion, Google in negotiations with Mechanize for a deal exceeding $1.5 billion, and OpenAI acquiring Ona, a cloud orchestration technology company. OpenAI’s Thibault Sottiaux, Core Products Lead, said of the Ona acquisition that it would “help us make Codex easier to deploy securely across production workflows,” per PYMNTS — indicating that infrastructure and orchestration technology, not just the coding-agent model itself, is becoming a competitive battleground.

This cluster of deals — Cognition’s talks, the SpaceX-Cursor figure, the Google-Mechanize negotiation, and OpenAI’s Ona acquisition — suggests investors and strategic acquirers are treating AI coding agents as infrastructure-critical rather than as a feature layer on top of existing developer tools. The valuations involved, from Cognition’s $40 billion target to Cursor’s reported $60 billion SpaceX deal, dwarf typical SaaS developer-tool valuations by an order of magnitude, reflecting a bet that AI coding agents will capture a meaningfully larger share of enterprise software spend than traditional IDE or CI/CD tooling ever did.

What This Means for Enterprise Buyers and Investors

1. Treat government/defense customer adoption as a credibility signal, not just a revenue line

Cognition’s addition of the U.S. Army and Navy to its customer list, alongside NASA, indicates the product has cleared security and procurement hurdles that are typically much higher than commercial enterprise sales. Enterprise buyers evaluating AI coding agents for regulated or security-sensitive environments should treat this as evidence the underlying security posture is mature enough for scrutiny, not just as a marketing data point.

2. Watch revenue-multiple compression risk if the $40B round closes

A valuation of $40 billion against an approaching-$1 billion revenue run rate implies a revenue multiple in the 40x range — high even by current AI-sector standards. Investors and later-stage employees evaluating equity exposure should model scenarios where growth decelerates from the current pace, since a multiple this rich has limited room for a growth slowdown before valuation pressure sets in.

3. Evaluate Devin’s positioning against “replacement” framing carefully

Scott Wu’s explicit statement that Devin is not sold as a human replacement is a deliberate market-positioning choice, not a technical limitation claim. Enterprise buyers should independently assess what tasks Devin can and cannot handle reliably, rather than taking either the “replacement” fear or the “augmentation only” reassurance at face value.

4. Track the OpenAI-Ona and Google-Mechanize deals as indicators of where the infrastructure layer is consolidating

The parallel deal activity around cloud orchestration (Ona) and coding-agent infrastructure (Mechanize) suggests the next competitive battle in AI coding tools may be fought at the infrastructure layer rather than the agent-model layer alone. Enterprises standardizing on a coding-agent vendor should ask about the underlying orchestration and deployment infrastructure roadmap, not just model capability.

The Correction Scenario

The bull case for Cognition’s valuation trajectory rests entirely on sustained triple-digit percentage revenue growth continuing at its current pace — a doubling roughly every three months is an extraordinary rate that very few companies sustain for more than a handful of quarters. If enterprise adoption growth decelerates even moderately — a plausible outcome once the initial wave of security-cleared, high-profile customers like the U.S. military and NASA is fully onboarded and usage growth normalizes to a steadier pace — a $40 billion valuation built on a 40x-plus revenue multiple would face real repricing pressure. The gap between Cognition’s current $492 million-to-nearly-$1 billion trajectory and the kind of growth rate needed to sustain a $40 billion mark over multiple years is the central risk any investor in this round is underwriting.

Frequently Asked Questions

Sources & Further Reading

• AI Coding Startup Cognition Reportedly Already in Talks to Raise at $40B Valuation — TechCrunch
• Cognition AI Eyes $40 Billion Valuation From New Funding — PYMNTS
• AI Coding Startup Cognition Reportedly in Talks for New Funding Round at $40B Valuation — The AI Insider