OpenAI Launches GPT-5.3 Codex, Boosting Codex with New Capabilities

OpenAI has launched GPT-5.3 Codex, an agentic coding model that powers Codex, while Anthropic has released its own competing model.

OpenAI Launches GPT-5.3 Codex, Boosting Codex with New Capabilities
OpenAI and Anthropic logos with AI model illustrations

On February 5, 2026, OpenAI unveiled GPT-5.3 Codex, an advanced agentic coding model designed to power Codex, the AI coding agent launched earlier. This model turns Codex from a simple tool for writing and reviewing code into an assistant capable of carrying out a wide range of professional tasks, expanding software development possibilities and improving work efficiency. According to OpenAI, GPT-5.3 Codex can build games and complex applications in a few days, at a speed 25% higher than the previous model, GPT-5.2. It is also the first model that helped create itself, as preliminary versions were used for debugging and performance evaluation. This launch came shortly after the release of a competing model by Anthropic, highlighting the rapid evolution in the field of agentic coding models. (techcrunch.com)

Comparison Between GPT-5.3 Codex and Anthropic's Model

The simultaneous release of advanced models by OpenAI and Anthropic underlines the intensifying competition in the field of AI applied to software development. Both models aim to improve the autonomy and efficiency of coding tools, but they differ in their capabilities and approaches. While GPT-5.3 Codex focuses on expanding Codex's features, enabling the creation of complex applications in less time, Anthropic's model stands out for its integration of advanced reasoning and memory capabilities, offering versatile performance across a range of computing applications, including financial tasks such as modeling and forecasting. (investing.com)

Implications for Software Development and the AI Industry

The introduction of these advanced models has significant implications for software development and the AI industry. Developers can expect more powerful and autonomous tools, capable of handling complex tasks with greater efficiency. However, the adoption of such technologies also raises questions about security, reliability and ethics in the use of AI. It is essential that companies and professionals in the sector carefully consider these aspects when integrating such tools into their workflows. (techradar.com)

Conclusion

  • Accelerated innovation: The simultaneous release of advanced models by OpenAI and Anthropic highlights the intensifying competition in AI applied to software development.
  • Potential impact: These models promise to improve the autonomy and efficiency of coding tools, with significant implications for software development and the AI industry.
  • Ethical considerations: Adopting these technologies raises questions about security, reliability and ethics in the use of AI, requiring careful evaluation by companies and professionals in the sector.