OpenAI, a prominent AI company, has unveiled GPT-5.3 Codex, a new model aimed at boosting the capabilities of its coding tool Codex. This launch comes shortly after a similar move by rival Anthropic, creating a stir in the tech community.
Initially introducing Codex, a coding tool, earlier this week, OpenAI quickly followed up with the release of GPT-5.3 Codex. This new model expands Codex’s functionality from writing and reviewing code to performing a wide range of tasks typically handled by developers and professionals on computers. OpenAI claims that GPT-5.3 Codex can expedite the development process, enabling the creation of complex games and applications from scratch within days.
Highlighting its efficiency, OpenAI states that GPT-5.3 Codex is 25% faster than its predecessor, GPT-5.2. Notably, this model stands out as the first one within the company that actively contributed to its own improvement. OpenAI’s staff utilized early versions of the program to debug and assess its performance, showcasing a unique evolution process.
The competitive dynamics in the tech landscape were further accentuated when Anthropic unexpectedly released its agentic coding model, narrowly edging out OpenAI in the timing race. Despite this, both companies have now introduced innovative solutions that promise to revolutionize coding practices and redefine software development norms.
Source: TechCrunch