Skip to content

VS Code 1.122: Chat still requires sign-in when using a configured OpenAI-compatible local network model in VSCode Portable Mode #319538

@BreathBlush

Description

@BreathBlush

Does this issue occur when all extensions are disabled?: Yes

  • VS Code Version: 1.122.1
  • OS Version: windows server 2019

Description

After updating to VS Code 1.122.1, the release notes indicate support for offline usage with local or LAN-hosted models.

I configured a custom OpenAI-compatible model endpoint and the model appears to work correctly. For example, sending a simple prompt such as hello returns a valid response.

However, when I ask the model to perform a Chat task (for example, "write a simple Python script"), VS Code shows a message indicating that I need to sign in before Chat can be used.

Steps to Reproduce:

  1. Install VS Code 1.122.1 (Windows x64 Portable).
  2. Enable portable mode by creating a data folder in the VS Code root directory.
  3. Configure a custom OpenAI-compatible model endpoint.
  4. Verify the model responds successfully to a simple prompt such as hello.
  5. Ask the model to generate a simple Python script.

Expected Behavior

The configured OpenAI-compatible local network model should be used for Chat requests without requiring a Microsoft/GitHub sign-in.

Actual Behavior

Simple prompts work, but Chat requests trigger a sign-in requirement instead of using the configured model.

Additional Notes

This is a restricted environment, so I am unable to provide screenshots. The issue is consistently reproducible with the setup described above.

Metadata

Metadata

Assignees

Labels

bugIssue identified by VS Code Team member as probable bugmodel-byoknew releaseIssues found in a recent release of VS Code

Type

No type
No fields configured for issues without a type.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions