CodeRabbit will conduct up to $5,000 worth of security scans per organization for free, for a total value of $1 million. In addition, the code review provider is joining OpenAI’s Daybreak Defense Network.
Deep Scan utilizes large language models (LLMs) such as OpenAI GPT-6.1 Sol. It assumes the role of a red team to thoroughly test an organization’s security. In addition to identifying vulnerabilities, the tool searches for genuinely exploitable cyber threats. The end result is a prioritized list with suggested fixes that align with the standards of the codebase.
CodeRabbit explains the generosity of the free scans as a key means of staying one step ahead of attackers, especially now that this group has gained extra speed thanks to AI. The company points to the much-discussed incidents of the past week. PHP urgently patched two critical vulnerabilities that had been in the code for 14 years, after which Cloudflare saw millions of exploitation attempts within a few hours. Microsoft reported that a state-affiliated group using an AI hacking tool attacked more than 2,250 cloud users.
“AI agents write more code than humans can review, creating new vulnerabilities,” says CEO Harjot Gill. “Moreover, anyone with a credit card can now exploit these vulnerabilities on a massive scale.”
8,000 serious vulnerabilities in one month
CodeRabbit Secure, of which Deep Scan is a part, found more than 8,000 critical vulnerabilities at over 600 organizations in the past 30 days. More than 70 percent stem from inadequate access controls, authentication errors, and SQL injections.
Traditional tools check code against fixed rules and known patterns, an approach that does not always detect these most common vulnerabilities. CodeRabbit, on the other hand, has multiple security agents simultaneously examine code paths, after which a semantic analysis determines whether a vulnerability can actually be exploited. This happens before, during, and after the merging of pull requests, ensuring a full lifecycle of secure code.
Integration with Daybreak
Through the Daybreak Defense Network, CodeRabbit aims to integrate OpenAI’s GPT cyber models more closely with CodeRabbit Secure. As part of the Daybreak initiative, OpenAI itself also offers Codex Security, which scans repositories, checks whether vulnerable code is accessible, and generates patches. This is very similar to what CodeRabbit offers.
Even outside of Daybreak, CodeRabbit faces significant competition, including from GitHub Copilot, GitLab Duo, Qodo, and Greptile. According to Sacra, the company raised $143 million in a Series C round this year, at a valuation of approximately $1.5 billion. CodeRabbit Secure is included in the Advanced and Enterprise plans, including the 14-day trial period for Advanced.
See also: OpenAI tackles open-source vulnerabilities with ‘Patch the Planet’