OpenAI launched GPT-5-Codex, an AI model for programming with greater speed, quality, and support on GitHub
Nuevo
Agregar La Derecha Diario en
Compartir:
OpenAI introduced its new GPT-5-Codex model, an artificial intelligence tool focused on programming. The announcement comes weeks after the company unveiled GPT-5 as a general model for ChatGPT.
The company aims to compete directly against solutions like Anthropic's Claude Code and Google's AlphaCode, targeting developers.
3. Resumir o corregir textos
What is GPT-5-Codex and how does it differ from GPT-5?
GPT-5-Codex is designed exclusively to generate and understand code in various programming languages. Unlike GPT-5, which is general-purpose, this version focuses on specific development tasks.
It can be accessed via CLI, web, integration with GitHub, IDEs, and also on mobile devices.
Main features and improvements
Produces higher-quality code with simple natural language instructions.
Adjusts response speed according to the complexity of the task.
Enables large-scale refactoring without altering code functionality.
Detects critical failures and runs automated tests to validate results.
Navigates complete codebases and analyzes dependencies.
GPT-5-Codex está diseñado exclusivamente para generar y entender código
Direct competition with other programming AIs
With GPT-5-Codex, OpenAI is entering the race to lead the programming assistant market. As highlighted in their official announcement, Codex usage grew tenfold in the past month.
The new model aims to position itself as the most reliable option compared to Anthropic and Google, who are also advancing their own code generation systems.
El uso de Codex creció diez veces en el último mes
Availability and access
To use GPT-5-Codex, users must have one of the ChatGPT subscriptions or access with free credits. Additionally, the company confirmed a deeper integration with GitHub for collaborative code reviews.
