OpenAI has just announced the latest of its models, GPT-5.4, and already, Microsoft has made it available on GitHub Copilot for developers to take advantage of. According to GitHub, its early testing revealed that GPT-5.4 “consistently hits new rates of success," with regards to agentic software development capabilities.
GPT-5.4 adds recent advances in reasoning, coding, and agentic workflows seen in the Codex series model to a general-purpose model. As a result of these improvements, GPT-5.4 can do complex work more accurately and efficiently than before. The model is notably better at deep web research queries and computer use.
GitHub says that the model shows improved logical reasoning and task execution when attempting to complete processes that require tools over multiple steps. The newest model will be available for Copilot Pro, Pro+, Business, and Enterprise users. In terms of availability, GitHub says you’ll be able to use the model in the following places:
- Visual Studio Code version v1.104.1 and later in all modes: chat, ask, edit, agent
- Visual Studio version 17.14.19 and later in all modes: agent, ask
- JetBrains versions 1.5.66 and later, in all modes: ask, edit, agent
- Xcode versions 0.48.0 and later in all modes: ask, agent
- Eclipse versions 0.15.1 and later in all modes: ask, agent
- github.com
- GitHub Mobile iOS and Android
- GitHub CLI
- GitHub Copilot Coding Agent
For organizations that want to enable GPT-5.4 usage, they will need to get their admins to enable the GPT-5.4 policy in Copilot settings. To learn more about this development, you can read GitHub's announcement here.
1 Comment
Load the comments and join the conversation!
Read the comments, ask the editors questions, show respect and join the conversation.