Introducing A-SWE: Powered by OpenAI's Agentic Software Engineer that writes, tests, and ships code autonomously.
"It's literally an agentic software engineer that can build an app for you. It can take a PR that you would give to any other engineer and go build it." — Sarah Friar, OpenAI CFO
Build entire applications independently, from start to finish with minimal human oversight. Takes requirements and turns them into production-ready code.
Processes pull requests like any human engineer would. Automatically reviews code, suggests improvements, and implements requested changes.
Performs thorough quality assurance, bug detection, and testing automatically during development - tasks software engineers typically dislike doing.
Creates comprehensive, up-to-date documentation for all code it writes - another task human developers often neglect or postpone.
Automatically detects, diagnoses, and fixes bugs in code. Performs comprehensive "bug bashing" to ensure application stability and reliability.
Powered by GPT-4.5's emotional intelligence enhancements, A-SWE excels at creative design tasks and understanding context beyond pure technical specifications.
Let A-SWE handle the routine tasks while your team tackles creative challenges.
A-SWE managing pull requests and conducting code reviews autonomously
OpenAI is developing A-SWE, an AI agent that can holistically replace software engineers by building apps and handling quality assurance, bugs and documentation.
"It's literally an agentic software engineer that can build an app for you. It can take a PR that you would give to any other engineer and go build it," says OpenAI CFO Sarah Friar.
OpenAI is set to launch A-SWE, an AI agent that can perform software engineering tasks along with quality assurance and bug testing.
A detailed comparison of OpenAI's A-SWE with other AI coding assistants, highlighting its autonomous capabilities that go far beyond code completion.
Read Article
Exploring how A-SWE fits into OpenAI's broader ecosystem alongside Deep Research and Operator, and what it means for the future of AI agents.
Read Article
What does the future hold for human developers as AI systems like A-SWE become more capable? An exploration of new roles and opportunities.
Read ArticleA-SWE (Agentic Software Engineer) is Powered by OpenAI's autonomous AI system designed to perform comprehensive software engineering tasks. Unlike GitHub Copilot or similar assistants, A-SWE can handle the entire development workflow from receiving requirements to shipping production-ready code.
Powered by OpenAI has announced that A-SWE is currently in development but has not specified an exact release date. Join our waitlist to be the first to know when it becomes available.
While tools like GitHub Copilot assist developers with suggestions and completions, A-SWE is designed to function as a complete software engineer. It can build entire applications, handle PRs, perform testing, fix bugs, and create documentation autonomously—tasks that currently require human engineers.
A-SWE is not designed to replace human developers but to augment development teams. It will likely shift human roles toward more strategic, creative, and oversight positions while handling routine coding tasks and quality assurance.
Join the waitlist to be notified when A-SWE launches