OpenAI rolls out GPT-5 Codex for coding automation
Summary
Hide ▲
Show ▼
OpenAI has begun rolling out the GPT-5 Codex model, an AI agent designed to automate coding tasks. The new model is integrated into various Codex instances, including Terminal, IDE extension, and Codex Web. It is intended to compete with Claude Code, the current market leader in AI coding. The GPT-5 Codex model achieves a 74.5% success rate on the SWE-bench Verified benchmark and shows significant improvement in code refactoring tasks. The rollout is expected to be completed in the coming days. The GPT-5 Codex model is designed to assist users in automating complex coding tasks, even for those who may not have extensive programming knowledge. This rollout aims to enhance the capabilities of OpenAI's coding tools, providing a more efficient and user-friendly experience for developers and non-developers alike.
Timeline
-
16.09.2025 03:51 1 articles · 13d ago
GPT-5 Codex model begins rolling out
OpenAI has started rolling out the GPT-5 Codex model to various Codex instances, including Terminal, IDE extension, and Codex Web. The model achieves a 74.5% success rate on the SWE-bench Verified benchmark and shows significant improvements in code refactoring tasks. The rollout is expected to be completed in the coming days.
Show sources
- OpenAI's new GPT-5 Codex model takes on Claude Code — www.bleepingcomputer.com — 16.09.2025 03:51
Information Snippets
-
GPT-5 Codex is an AI agent designed to automate coding tasks.
First reported: 16.09.2025 03:511 source, 1 articleShow sources
- OpenAI's new GPT-5 Codex model takes on Claude Code — www.bleepingcomputer.com — 16.09.2025 03:51
-
The model is integrated into Terminal, IDE extension, and Codex Web.
First reported: 16.09.2025 03:511 source, 1 articleShow sources
- OpenAI's new GPT-5 Codex model takes on Claude Code — www.bleepingcomputer.com — 16.09.2025 03:51
-
The GPT-5 Codex model achieves a 74.5% success rate on the SWE-bench Verified benchmark.
First reported: 16.09.2025 03:511 source, 1 articleShow sources
- OpenAI's new GPT-5 Codex model takes on Claude Code — www.bleepingcomputer.com — 16.09.2025 03:51
-
The model shows a 51.3% success rate in code refactoring evaluations, up from 33.9% with the previous GPT-5 model.
First reported: 16.09.2025 03:511 source, 1 articleShow sources
- OpenAI's new GPT-5 Codex model takes on Claude Code — www.bleepingcomputer.com — 16.09.2025 03:51
-
The rollout is expected to be completed in the coming days.
First reported: 16.09.2025 03:511 source, 1 articleShow sources
- OpenAI's new GPT-5 Codex model takes on Claude Code — www.bleepingcomputer.com — 16.09.2025 03:51