fix: use LM Studio native API endpoint (/api/v1) instead of OpenAI-compatible (/v1)

[roomote-v0]

Related GitHub Issue

Closes: #12223

Description

This PR attempts to address Issue #12223. The LM Studio provider was hardcoding the OpenAI-compatible /v1 path for all API requests, even though the user selected "LM Studio" as the provider. This meant requests went to /v1/chat/completions and /v1/models instead of the native LM Studio endpoints at /api/v1/chat/completions and /api/v1/models.

The native /api/v1 endpoint (introduced in LM Studio 0.4.0) provides features like stateful chats, prompt caching, and context length management that are not available through the OpenAI-compatible endpoint.

Changes made:

• src/api/providers/lm-studio.ts: Changed OpenAI client baseURL from /v1 to /api/v1, and model listing endpoint from /v1/models to /api/v1/models
• src/api/providers/fetchers/lmstudio.ts: Changed connection test and model fetching endpoints from /v1/models to /api/v1/models
• Updated all corresponding test assertions to expect the new /api/v1 paths

Feedback and guidance are welcome.

Test Procedure

• All existing LM Studio tests updated and passing (24/24)
• Run: cd src && npx vitest run api/providers/__tests__/lmstudio.spec.ts api/providers/__tests__/lm-studio-timeout.spec.ts api/providers/fetchers/__tests__/lmstudio.test.ts
• Manual verification: With LM Studio running locally, selecting "LM Studio" as provider should now send requests to /api/v1/* endpoints instead of /v1/*

Pre-Submission Checklist

• Issue Linked: This PR is linked to an approved GitHub Issue (see "Related GitHub Issue" above).
• Scope: My changes are focused on the linked issue (one major feature/fix per PR).
• Self-Review: I have performed a thorough self-review of my code.
• Testing: New and/or updated tests have been added to cover my changes (if applicable).
• Documentation Impact: No documentation updates required -- this is an internal endpoint path change.
• Contribution Guidelines: I have read and agree to the Contributor Guidelines.

Documentation Updates

• No documentation updates are required.

Additional Notes

The native LM Studio API at /api/v1 uses the same response format as the OpenAI-compatible endpoint, so the OpenAI SDK client continues to work without any parsing changes. The key benefit is that the native endpoint enables LM Studio-specific features like stateful chats and prompt caching.

Interactively review PR in Roo Code Cloud