Skip to main content

VS Code Custom Model Clients with Keeptrusts

This integration applies only when the client exposes an API base URL or a complete model endpoint. A generic network proxy setting is not equivalent.

Required client capabilities​

Before configuration, confirm that the client can set:

  • the endpoint or OpenAI-compatible base URL;
  • the model ID;
  • a bearer/API-key credential;
  • the API family it sends, such as Chat Completions or Responses;
  • any tool-calling or streaming capabilities the selected feature requires.

If any of these are fixed to the extension vendor, that traffic cannot be routed through the Keeptrusts model API.

Gateway values​

Client fieldValue
OpenAI-compatible base URLhttp://127.0.0.1:41002/v1
Full Chat Completions URLhttp://127.0.0.1:41002/v1/chat/completions
Full Responses URLhttp://127.0.0.1:41002/v1/responses
API keyA Keeptrusts client API token
ModelAn ID returned by /v1/models

Use only the URL shape the client asks for; supplying /v1 where it expects a full route can produce a misleading 404.

VS Code also has a built-in Custom Endpoint provider under Chat: Manage Language Models. Its current setup is covered in VS Code and GitHub Copilot.

Verify​

Select the custom model, run kt events tail --since 10m --follow, and send one small request. Do not claim the integration covers inline suggestions, utility models, embeddings, or extension telemetry until each of those calls appears in gateway events.

Next steps​