GPT-5.3-Codex is OpenAI’s most advanced agentic coding model. It pairs the frontier software engineering performance of GPT-5.2-Codex with the broader reasoning and professional knowledge capabilities of GPT-5.2. It delivers state-of-the-art results on SWE-Bench Pro and strong performance on Terminal-Bench 2.0 and OSWorld-Verified, highlighting better multi-language coding, terminal fluency, and real-world computer-use skills. The model is tuned for long-running, tool-driven workflows and supports interactive steering during execution, making it well-suited for complex development work, debugging, deployment, and iterative product cycles. Outside of coding, GPT-5.3-Codex also performs well on structured knowledge-work benchmarks such as GDPval, enabling tasks like drafting documents, analyzing spreadsheets, creating slides, and conducting operational research across domains. It is trained with increased cybersecurity awareness, including the ability to identify vulnerabilities, and is deployed with extra safeguards for higher-risk scenarios. Relative to earlier Codex models, it is more token-efficient and about 25% faster, aimed at end-to-end professional workflows that combine reasoning, execution, and computer interaction.
Not enough uptime data to display a chart