OpenAI is launching a new proxy cryptography model just minutes after Anthropic dropped its own model

OpenAI on Monday launched Codex, an agent programming tool marketed to software developers. Today, OpenAI also launched a new model designed to drive Codex: GPT-5.3 Codex.

The company says the model transforms Codex from an agent that can only “write and review code” to an agent that can do “almost anything developers and professionals do on a PC, expanding who can create software and how work gets done.” After testing its new model against a number of performance benchmarks, OpenAI claims it can create “complex, high-performance games and applications from scratch over the course of days.”

OpenAI says that the GPT-5.3 Codex is also 25% faster than its previous model (GPT-5.2), and that it was the company’s first model “that was instrumental in creating itself,” meaning company employees used early versions of the software to debug and evaluate how it performed.

The release of the new model comes notably on the heels of a new agent programming model released by its competitor Anthropic. In fact, OpenAI and Anthropic originally planned to release the two proxy encryption tools at the exact same time: 10 a.m. PT. However, shortly before the original release time, Anthropic moved up its release date by 15 minutes, slightly beating OpenAI in the race to publish models.