Runars DocsCursor
Point Cursor at Runars via its OpenAI-compatible custom model
Cursor is an AI code editor. Its "custom model" feature lets you route requests to any OpenAI-compatible endpoint—Runars's /v1/chat/completions is that endpoint.
Prerequisites
- A Runars API key (Dashboard → API Keys), starting with
sk-runars- - A Cursor Pro / Business account (custom models are a paid feature)
- Credit balance on your organization
Setup
Open Settings → Models, then toggle on Custom Model (or open the model picker and choose "Custom Model").
Configure it:
| Field | Value |
|---|---|
| Provider | OpenAI |
| Base URL | https://api.runars.ca/v1 |
| API Key | sk-runars-... |
| Model id | glm-5.2 (or any id from GET /v1/models) |
Save, then click Verify. A green checkmark means Cursor can reach Runars.
Using it
Pick your custom model from the model dropdown in Chat or Cmd-K. Every request Cursor makes flows through Runars and is debited against your organization's credit balance.
Troubleshooting
"Model not found": Cursor may send a model id that isn't in your Runars catalog. Open the Cursor model settings and set the model id explicitly to one returned by GET /v1/models (e.g. glm-5.2).
"401 Unauthorized": The API key is missing, wrong, or revoked. Keys are cache-invalidated within ~5s of revocation on Runars.
Verify button fails: Confirm the Base URL ends at /v1 (not /v1/chat/completions — Cursor appends the path). Confirm the key is exported and copied without trailing whitespace.
What's next
- Install SDKs — Anthropic & OpenAI SDK setup
- Pricing & billing — credit model, rate limits, auto-reload