Microsoft Launches GPT-5-Codex in Azure AI Foundry
The model will offer a major boost to developers' workflow
After announcing the launch of GPT-5 Codex earlier this month, Microsoft has now rolled out the coder-friendly model to Azure AI Foundry. Besides, the company has also announced that GPT-5-Codex is available in public preview for GitHub Copilot in Visual Studio Code.
Unlike earlier Codex models, GPT-5-Codex accepts both text and image prompts in a single workflow. That means developers can handle complex, repository-scale tasks without switching between multiple tools.
Microsoft says this will help teams deliver context-aware solutions faster while staying in the flow of development. Not to forget, the model is also designed for practical integration. Meaning, developers can use the Response API in Azure AI Foundry to automate coding tasks, run tests, and shorten review cycles.
That’s not all; GPT-5-Codex further brings built-in expertise for code reviews, too. It details hidden issues early to raise code quality without the usual manual bottlenecks.
Microsoft highlighted several real-world use cases in the announcement blog post. Teams can rely on GPT-5-Codex as a repo-aware refactoring assistant, a flaky test that loops until stability, or even a cloud migration copilot that edits scripts and handles CLI errors intelligently.
On the pricing front, GPT-5-Codex follows a pay-per-million-tokens model. Input is billed at $1.25, cached input at $0.125, and output at $10. Deployment starts with East US 2 and Sweden Central regions, with global availability coming soon.
All in all, the availability of GPT-5 Codex in Azure AI Foundry ensures advanced automation, multimodal inputs, and code review capabilities into one streamlined experience.
Read our disclosure page to find out how can you help Windows Report sustain the editorial team. Read more
User forum
0 messages