Cognition is in talks to value Devin at $40bn, three months after $26bn — AI coding agents keep re-pricing what 'AI can code' means

Cognition, maker of the AI coding agent Devin, is reportedly in early talks to raise at a $40bn+ valuation — up more than 50% from the $26bn it raised at just three months ago — with annualised revenue approaching $1bn, the latest data point in investors betting AI agents can do real, unsupervised engineering work.

Cognition, the company behind the autonomous coding agent Devin, is reportedly in early talks with investors for a new funding round that could value it at $40bn or more — up over 50% from the $26bn valuation it raised at just three months earlier, when it took in $1bn. Behind the number: annualised revenue run rate has roughly doubled since that last raise, closing in on $1bn. The round, and its size, are still early and unconfirmed, but the trajectory itself is the story — three valuation step-changes inside a year for a company selling the premise that an AI agent can read a codebase, plan the work, write the code, run the tests, and fix its own mistakes with minimal human oversight.

This isn’t an isolated data point. It sits alongside Cursor’s climb from $1bn to $2bn ARR in three months, Lovable’s $13.3bn valuation on a $400m raise, and a broader AI-coding-startup deal-size trend that’s gone from an average $7.4m in 2022 to $527.8m by the end of 2025. Investors aren’t pricing these companies on “helpful autocomplete” anymore — they’re pricing them on the bet that autonomous agents genuinely substitute for engineering hours at scale, and doing it with real, growing revenue behind the bet rather than speculation alone.

So what

The valuations are a signal worth reading, not a reason to hand a production codebase to an unsupervised agent. Every serious team using tools like Devin, Claude Code or Cursor in production still puts human review, testing discipline and architectural judgement around the output — the capability gap between “drafts working code fast” and “ships production software safely” hasn’t closed just because the funding round did. If you’re weighing how much of your build to hand to AI-driven tooling versus experienced engineers who know where that line sits, our AI-assisted development work is built around getting that balance right. Get in touch to talk through what it looks like for your project.

← All research

More research.

MAI Code 2 and MAI Code 2 Flash arrive in GitHub Copilot, including a model for free-plan users

GitHub has added Microsoft's MAI Code 2 and MAI Code 2 Flash to Copilot across editors, the CLI, the cloud agent and mobile, with administrators able to control access through model policy.

Read ›

GitHub Copilot team customizations are generally available: set default agent plugins for a whole team

GitHub has made team customizations generally available, so team maintainers and organisation owners can set approved Copilot agent plugins as defaults for every member of a team.

Read ›

Claude Code 2.1.330: scheduled tasks run for 30 days, risky rm commands ask first, and enterprise sign-in arrives

Claude Code 2.1.330 extends scheduled tasks and /loop from 7 to 30 days, makes more destructive rm commands ask for approval, and adds sign-in for organisations managed in Anthropic's Enterprise Admin Console.

Read ›

OpenAI's GPT-6.2 Astra and Sol are generally available in GitHub Copilot, with a code review focus

GitHub made OpenAI's GPT-6.2 Astra and GPT-6.2 Sol generally available in Copilot for agentic coding and code review. GitHub reports early tests catching more real issues.

Read ›

Copilot code review adds adaptive effort and secret detection in public preview

GitHub's Copilot code review can now choose Lite, Balanced or Max effort per pull request and flag exposed secrets, both in public preview. Azure DevOps support is generally available.

Read ›

Claude Code 2.1.329: /code-review now covers the whole branch, and saved scheduled tasks need rescheduling

Claude Code 2.1.329 makes /code-review review the whole branch against the default branch and adds fixes for Code Mode and sign-in. Saved scheduled tasks mostly need setting up again.

Read ›

Start a project

Building with AI? Let's talk.

Tell us what you are building. We'll give you an honest view of the right approach.

Start a projectSee our work