# Cursor: Claude and Codex configuration

> Use Cursor's custom OpenAI-compatible model fields for either ScaleMax model family.

> **Cursor support is experimental.** Cursor does not publish a stable external configuration contract for every custom-model feature. A Cursor update can rename settings, reset a custom model, or route some editor-only features through Cursor's own service.

## Automatic configuration

**macOS / Linux**

```bash
npm install -g scalemax@latest
scalemax
```

**Windows PowerShell**

```powershell
npm.cmd install -g scalemax@latest
scalemax.cmd
```

Windows PowerShell blocks npm’s .ps1 launchers by default (“running scripts is disabled on this system”), so these commands use the .cmd launcher: npx.cmd, npm.cmd, scalemax.cmd. In Command Prompt you can leave out .cmd.

Select **Cursor**. The CLI makes a backup before updating the local Cursor profile, includes every model returned for the key (including a Claude-only catalog through Cursor's OpenAI-compatible route), and asks you to restart Cursor. Run `scalemax doctor` after every Cursor upgrade. If verification fails, use the manual fields below; do not repeatedly patch an unknown Cursor version.

## Manual configuration for Codex models in Cursor

Open **Cursor Settings → Models**, add a custom OpenAI-compatible model, and enter:

- **API-credit base URL:** `https://api.scalemax.pro/v1`
- **OpenAI token-pack base URL:** `https://api.scalemax.pro/token/v1`
- **API key:** the full key that belongs to that URL
- **Model:** one exact Codex model ID below

- `gpt-6.1-sol`
- `gpt-6-sol`
- `gpt-6-luna`
- `gpt-6-astra`
- `gpt-5.6-terra`
- `gpt-5.6-sol`
- `gpt-5.6-luna`

Cursor's custom provider uses its OpenAI-compatible chat path; it is not the native Codex Responses CLI. Features that depend on Cursor's own model contracts may remain unavailable even when chat and agent requests succeed.

## Manual configuration for Claude models in Cursor

Use the same **custom OpenAI-compatible** form, not Cursor's built-in Anthropic account login:

- **API-credit base URL:** `https://api.scalemax.pro/v1`
- **Claude token-pack base URL:** `https://api.scalemax.pro/token/v1`
- **API key:** the full key that belongs to that URL
- **Model:** one exact Claude model ID below

- `fable-5`
- `claude-haiku-4-5-20251001`
- `claude-haiku-5-5`
- `claude-opus-4-6`
- `claude-opus-4-7`
- `claude-opus-4-8`
- `claude-opus-5`
- `claude-opus-5-5`
- `claude-sonnet-5`
- `claude-sonnet-5-5`

This is Claude through Cursor's OpenAI-compatible adapter. It is not the native Anthropic Messages setup used by Claude Code.

## Verify without a large request

Create a new empty chat, select the custom ScaleMax model, disable repository context, and send `Reply with K only.` A single-letter answer confirms the key, URL, and selected model. Then test one read-only agent action before trusting write or terminal operations.

## Key and feature boundaries

> Do not reuse a key on the other billing URL, paste it into a workspace file, or share Cursor's local profile. ScaleMax can validate the upstream model request, but it cannot guarantee Cursor-owned autocomplete, background indexing, or proprietary editor features.