CodeRabbit and Greptile Deploy GPT-6.1 Sol and GPT-6 Astra for Code QA

AI developer tooling startups CodeRabbit and Greptile announced new production deployments on Monday built on OpenAI’s GPT-6.1 Sol and GPT-6 Astra models, extending automated pull-request review into cross-repository vulnerability validation and live browser-based application testing, according to OpenAI case studies.

CodeRabbit Secure Uses GPT-6.1 Sol for Cross-Repo Deep Scans and Sandbox Verification

CodeRabbit is using GPT-6.1 Sol as the primary reasoning model inside its security agent to trace how a pull request affects shared libraries, dependencies, and authentication safeguards across complex codebases. While standard pull-request review analyzes changes as developers commit code, CodeRabbit’s Deep Scan and CodeRabbit Secure extend analysis across linked repositories—such as shared authentication packages—across thousands of organizations.

For user-selected findings, CodeRabbit Secure can attempt exploit verification inside an isolated sandbox without touching production systems, and can draft a pull request containing a proposed fix for human engineers to review and approve.

Greptile Pairs GPT-6 Astra Computer Use With Daybreak and Codex MCP

Meanwhile, Greptile co-founder and CEO Daksh Gupta said the company is using GPT-6 Astra’s computer-use and visual capabilities to automate live application testing at scale. For each pull request, Greptile can spin up a development server, resize a browser to an iPhone viewport, and simulate user interactions to catch UI regressions—such as a disappearing button—that static code reading would miss.

Greptile also highlighted Daybreak, a defensive security testing service for startups and enterprises, and reported that thousands of customers have connected OpenAI’s Codex to Greptile’s codebase Knowledge Base via a Model Context Protocol (MCP) server, reducing the rate at which AI-generated code changes are reverted.