Skip to content

Issue Title: LM Studio - Initial Connection Test Doesn't Check for unloaded Models #306

@haiphucnguyen

Description

@haiphucnguyen

Steps to Reproduce:

  1. Start LM Studio on Windows x64.
  2. Open the Askimo interface within LM Studio.
  3. Navigate to Settings -> AI Provider Configuration.
  4. Select “LM Studio” as the provider.
  5. Accept the default connection setting: http://localhost:1234/v1
  6. Click the “Test Connection” button.
  7. Observe: "No models available for lmstudio" error is displayed.

Expected Result:
The “Test Connection” button should attempt to establish a connection with the LM Studio API at http://localhost:1234/v1 without requiring a model to be loaded within LM Studio. It should indicate whether the connection is successful, regardless of whether models are currently loaded or not.

Actual Result:
The test connection fails with an error message indicating no models are available, even though LM Studio is running and the API should be accessible. This leads to a confusing user experience, as it suggests that a model must be loaded before testing the connection.
Suggested Solution:
Modify the “Test Connection” functionality to:
• Perform a simple HTTP connection test to http://localhost:1234/v1 without requiring a model to be loaded.
• Return an appropriate success or failure message based on the connection status, regardless of whether any models are loaded.

Additional Notes:
As noted by the user, this behavior is counterintuitive and suggests a need for clearer communication about the connection process. A more direct test of API connectivity would improve usability and reduce user confusion.

Metadata

Metadata

Assignees

No one assigned

    Labels

    help wantedExtra attention is needed

    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