About OpenAI o3 in GitHub Copilot Chat
Note
o3 in Copilot Chat is currently in 公共预览版 and subject to change.
OpenAI has a family of large language models that you can use as an alternative to the default model used by Copilot Chat. o3 is one of those models and excels at coding tasks across the entire software development lifecycle, from initial design to bug fixes, maintenance to optimizations. For information about the capabilities of o3, see the OpenAI documentation.
o3 is currently available in:
- Copilot Chat in Visual Studio Code
- Immersive mode in Copilot Chat in GitHub
o3 is hosted by OpenAI and GitHub's Azure tenant when used in GitHub Copilot. OpenAI makes the following data commitment: We [OpenAI] do not train our models on your business data by default. GitHub maintains a zero data retention agreement with OpenAI.
When using o3, input prompts and output completions continue to run through GitHub Copilot's content filters for public code matching, when applied, along with those for harmful, offensive, or off-topic content.
Configuring access
You must enable access to OpenAI o3 individually before you can use the model.
Setup for individual use
If you have a Copilot Free or Copilot Pro subscription, you can enable OpenAI o3 in two ways:
-
The first time you choose to use o3 with Copilot Chat in Visual Studio Code, or in the immersive view of Copilot Chat, you will be prompted to allow access to the model.
Clicking Allow enables you to use o3 and updates the policy in your personal settings on GitHub.
-
You can enable the model directly in your personal settings on the GitHub website. See 以单个订阅者身份管理 Copilot 策略.
Setup for organization use
As an organization owner, you can enable or disable o3 for everyone who has been assigned a Copilot Business seat through your organization. See 管理组织中的 Copilot 策略.
Using o3
For details of how to change the model that Copilot Chat uses, see 正在更改 Copilot Chat 的 AI 模型.