fix(vllm): support LM Studio endpoints

[waleedlatif1]

Summary

• pass local provider environment variables into the Ollama Compose app service
• accept both server-root and versioned OpenAI-compatible URLs for vLLM and LM Studio
• use the shared Chat Completions token-limit parameter and document the supported setup

Type of Change

• Bug fix

Testing

• 24 focused provider and model-discovery tests
• repository lint and type-check
• 33 CI audits and block-registry validation
• Docker Compose configuration validation

Checklist

• Code follows project style guidelines
• Self-reviewed my changes
• Tests added/updated and passing
• No new warnings introduced
• I confirm that I have read and agree to the terms outlined in the Contributor License Agreement (CLA)

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

1 Skipped Deployment

Project	Deployment	Actions	Updated (UTC)
docs	Skipped		Aug 24, 2026 5:20pm

[cursor]

PR Summary

Low Risk
Changes are localized to the vLLM/OpenAI-compat provider path and self-host docs; token parameter rename may affect vLLM servers that only accepted max_completion_tokens, though that’s the intended LM Studio fix.

Overview
Adds LM Studio as a documented self-hosted option alongside vLLM, using the existing VLLM_BASE_URL / VLLM_API_KEY settings and the same vllm/ model-picker prefix.

Runtime: Introduces getOpenAICompatibleApiBaseUrl so discovery and chat calls accept either a server root or a URL that already ends in /v1, without double-appending /v1. Invalid bases with query strings or fragments yield an empty model list instead of a bad fetch. The vLLM provider now sends max_tokens (not max_completion_tokens) for token limits, matching LM Studio’s Chat Completions API.

Docker: docker-compose.ollama.yml loads an optional .env on simstudio so VLLM_* (and other local provider vars) reach the container without inline overrides.

Docs cover LM Studio setup, env var wording, and troubleshooting when requests wrongly route to Ollama if models aren’t discovered via the vllm/ prefix.

^{Reviewed by Cursor Bugbot for commit e5a7cdc. Configure here.}

[greptile-apps]

Greptile Summary

This PR adds LM Studio support to the existing vLLM integration and completes the fixes requested in the previous URL-handling threads.

• Normalizes server-root and already-versioned OpenAI-compatible endpoint URLs through a shared helper.
• Keeps URL normalization within graceful model-discovery and provider-initialization error boundaries.
• Uses the shared Chat Completions token-limit parameter and forwards local-provider environment configuration through Docker Compose.
• Adds focused tests and self-hosting documentation for LM Studio setup.

Confidence Score: 5/5

The PR appears safe to merge.

No blocking failure remains; the current code fixes both previously reported URL-handling issues by parsing pathname components safely and keeping normalization within the graceful models-route error boundary.

Important Files Changed

Filename	Overview
apps/sim/providers/openai-compat/base-url.ts	Parses endpoint URLs, rejects query and fragment components, and normalizes only the pathname to a single /v1 API base.
apps/sim/app/api/providers/vllm/models/route.ts	Uses shared URL normalization within the existing graceful error boundary before requesting the models endpoint.
apps/sim/providers/vllm/index.ts	Reuses normalized API bases for discovery and inference while preserving endpoint validation and pinned-fetch behavior.
docker-compose.ollama.yml	Optionally loads the adjacent .env file into the Sim application service so local-provider configuration reaches the container.

_{Reviews (3): Last reviewed commit: "fix(vllm): guard discovery URL validatio..." | Re-trigger Greptile}

[waleedlatif1]

@greptile

[waleedlatif1]

@cursor review

[waleedlatif1]

@greptile

[waleedlatif1]

@cursor review

[cursor]

Cursor Bugbot has reviewed your changes using high effort and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 1505233. Configure here.}

[cursor]

✅ Bugbot reviewed your changes and found no new issues!

Comment @cursor review or bugbot run to trigger another review on this PR

^{Reviewed by Cursor Bugbot for commit e5a7cdc. Configure here.}