CodeRabbit raises $143M at $6B valuation as AI code review becomes the new observability

CodeRabbit raises $143M at $6B valuation

CodeRabbit, an AI code review platform founded two and a half years ago, has closed a $143M Series C at a $6B valuation. Atomico led the round, with BMW Ventures and Datadog participating. BMW was an early customer before becoming an investor. The $6B figure implies roughly $500M of value created per year since founding.

The growth driver

Gill's thesis is straightforward: AI code generation is the biggest use case in software today, and it has created a downstream bottleneck at code review. As Claude Code, Codex, and similar tools push more code into pull requests faster, the validation step becomes the chokepoint. CodeRabbit sits there.

The addressable pool is also expanding. Product managers and designers with no traditional coding background are now opening pull requests. Companies like Spotify, Notion, Linear, and Datadog are shipping coding products of their own. Every organization generating more code needs somewhere for that code to land and be validated before it ships.

“We just ended up raising $143,000,000 at a $6,000,000,000 valuation. We started two and a half years back. The role of human has become more of a reviewer now — you're waking up every day deciding which AI task to take to completion, where to add your taste. That's what we're helping scale.”

What CodeRabbit actually does

Gill describes three distinct functions. First, automated pre-merge checks, essentially pitting Claude Code against CodeRabbit in a loop to enforce guardrails before anything merges. Second, triage and prioritization for human reviewers, helping them decide which AI-generated tasks are worth their attention. Third, and arguably the most interesting: raising the abstraction layer so reviewers understand the blast radius and architectural impact of changes without reading individual lines of code.

His observation that even he is no longer reading code is a signal of how far the inflection has moved in the last three to six months.

The investor skepticism

The fundraise was not frictionless. Gill says investors are anxious about the entire software development lifecycle space, with capital concentrating heavily on model labs like Anthropic. The deeper question was whether code review as a discrete step survives at all, especially as figures like Dario Amodei argue software engineers may no longer be necessary.

Gill's answer draws on his earlier career in observability. When microservices and cloud infrastructure created chaos for operators in the early Datadog era, a new category formed around making complex systems legible. Code review, he argues, is the same transition: AI-generated code creates an explainability problem that cannot be skipped, and GitHub is the one place where changes must be inspected and approved before reaching production. The analogy landed with Datadog itself, which joined the round as an investor.

On the Cursor comparison, Gill notes that code editors have already proven fragile. Cursor built around tab completion and got outflanked by Claude Code, which he compares to Google displacing Yahoo. Code review guardrails, by contrast, are structurally upstream of whatever IDE or agent is generating the code. They do not get displaced by a model upgrade.

International footprint

CodeRabbit says it saw organic international traction early, including going viral in Japan shortly after launch. Atomico's involvement as lead investor reflects genuine European pull rather than a deliberate expansion push, and BMW's progression from customer to investor suggests similar dynamics in enterprise markets outside the US.