Connect Cursor to OpenLLM Buddy

If you do not already have Cursor, install the desktop app for macOS, Windows, or Linux from the official site.

1. Download Cursor from cursor.com.
2. Sign in or create a Cursor account when prompted.
3. Open a project folder so Chat and Agent have workspace context.

Cursor sends your OpenLLM Buddy API key as a Bearer token when you use your own OpenAI-compatible endpoint. Create the key before changing model settings.

1. Open the API keys page in the OpenLLM Buddy console.
2. Create a key and copy it immediately.
3. Ensure you have an active deployment for the model you plan to use.

You will paste this key into Cursor's OpenAI API Key field in the next steps — not your Cursor subscription key.

Model and API overrides live in Cursor Settings (the Cursor app settings), not the generic VS Code settings JSON.

1. Open Cursor Settings — click the gear in the top-right, or use Cursor → Settings → Cursor Settings on macOS (Windows/Linux: File → Preferences → Cursor Settings).
2. Go to the Models section in the sidebar.
3. Under OpenAI API Key, paste your OpenLLM Buddy API key and enable verification if Cursor prompts you.

UI labels can vary slightly by Cursor version. See Cursor API key documentation if your layout differs.

Turn on the override so Cursor sends OpenAI-format requests to OpenLLM Buddy instead of OpenAI's servers.

1. In Models, find Override OpenAI Base URL (or similar) and enable it.
2. Set the URL to your OpenLLM Buddy API root ending in /v1 — not a dashboard link and not a /chat/completions path.
3. Save or confirm when Cursor validates the key against that base URL.

• Base URL (example)https://openllmbuddy-proxy.botbuddytech.workers.dev/v1
• API Key (example)YOUR_API_KEY

If validation fails, run the curl test in step 7 with the same Base URL and key before retrying in Cursor.

Cursor needs the exact model handle your OpenLLM Buddy deployment exposes over the Chat Completions API.

1. In Models, use Add model (or the custom model field) if Cursor does not list your handle yet.
2. Enter qwen3.6:27bas the model name — matching this guide's target model.
3. Confirm the model appears in the model list and is enabled for Chat / Agent as needed.

Other handles include gemma4:26b and qwen3.6:27b. See available models for the full list tied to your templates.

After saving settings, pick your custom model from the model dropdown in the Chat sidebar or inline composer.

1. Open the Chat panel (Cmd+L / Ctrl+L).
2. Open the model picker and select qwen3.6:27b (or the name you added).
3. Send a short prompt to confirm you get a response from your deployment.

Composer and Agent modes use the same model list when configured for your own API key. Usage billing for BYOK goes to your OpenLLM Buddy deployment, not Cursor's bundled models.

Before relying on Cursor, confirm your API key, Base URL, and model handle work with a direct Chat Completions request.

curl -sS "https://openllmbuddy-proxy.botbuddytech.workers.dev/v1/chat/completions" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen3.6:27b",
    "messages": [{"role": "user", "content": "Say hello in one sentence."}],
    "max_tokens": 64
  }'

You should receive JSON with an assistant message. 401 means a bad API key — check the console. 404 usually means the Base URL includes /chat/completions — use only the /v1 root in Cursor.

• Invalid API key — use an OpenLLM Buddy key from the console, not a Cursor subscription token.
• Wrong Base URL — must end with /v1. Remove /chat/completions from the override field.
• Model not found — the handle in Cursor must match your deployment (e.g. gemma4:26b). Redeploy or pick a model from models.
• Still on OpenAI models — select your custom model in the Chat dropdown; bundled Cursor models ignore your override.

If your Cursor build does not show Base URL override, use the Cline extension path in the section below — same OpenAI-compatible wiring as our VS Code guide.

You are set. Cursor Chat and Agent should now call your OpenLLM Buddy deployment using your API key and model handle. Monitor usage and uptime from the console.

Explore more guides on the integrations page.

Cursor is VS Code–compatible. If you prefer an extension-based agent sidebar (or your Cursor build lacks Base URL override), install Cline inside Cursor and follow our VS Code guide — the steps are the same: OpenAI Compatible provider, Base URL, API key, and Model ID.

Open the VS Code + Cline guide →

Use this path only when native Cursor model settings are unavailable. The native override in steps 3–6 is the recommended approach for Cursor IDE.

Quick reference for Cursor + OpenLLM Buddy wiring.

Settings location

Cursor Settings → Models

Override Base URL

https://openllmbuddy-proxy.botbuddytech.workers.dev/v1

Model ID

qwen3.6:27b

Chat Completions URL

https://openllmbuddy-proxy.botbuddytech.workers.dev/v1/chat/completions

Auth header

Authorization: Bearer YOUR_API_KEY

Notes

OpenLLM Buddy uses Chat Completions. Paste the /v1 root in Cursor — not the full completions path.