---
title: Cursor compatibility and configuration
sidebarTitle: Cursor
description: Find Cursor's custom API settings and understand what still needs to be verified before using Infistar.
---

Cursor supports a custom OpenAI-compatible base URL. Whether a particular Infistar model works depends on your Cursor version, account permissions and the API format required by that model.

## Settings

Cursor's official support instructions describe **Settings → Models → API Keys**, with **Use OpenAI API Key** and **Override OpenAI Base URL**. A team administrator can restrict personal API keys or hide these controls.

1. Check that your account permits personal API keys.
2. Enter your own Infistar key and use `https://infistar.ai/v1` for international access, or `https://infistar.cc/v1` for mainland China.
3. Copy the exact model ID from models and pricing. Enable the key and custom URL, then start a new local chat to test.
4. Confirm that requests reach Infistar and verify the model's required API format, streaming and tool support.

## Verification boundary

We have checked that Cursor offers these settings. We have not completed end-to-end testing of the current Cursor version with every Infistar model. A configurable endpoint is not a guarantee that every model or feature works.

Bringing your own key does not replace all Cursor features or charges. Features such as Tab completion have separate restrictions; check Cursor's current documentation and your account settings.

Sources checked on 2026-10-21: [Cursor API Keys](https://docs.cursor.com/settings/api-keys) and [Cursor's official support discussion of custom endpoints and team permissions](https://forum.cursor.com/t/how-to-adjust-settings-for-custom-added-models/173253).

For existing terminal setup guides, see [Codex](/en/client-integrations/codex) and [Claude Code](/en/client-integrations/claude-code). Do not expose your API key or use unknown proxies to work around a client restriction.