IDE Integration

Overview

AIone supports two integration methods:

OpenAI-compatible format: For Cursor, Cline, Continue, and most IDEs or plugins that support a custom OpenAI Base URL

Anthropic native format: For Claude Code, the Anthropic SDK, and tools that call the Messages API directly

This guide covers how to configure AIone in popular IDEs and coding tools.

Cursor

Cursor is one of the most popular AI code editors.

Setup

Open Cursor > Settings > Models

Find the OpenAI API Key configuration

Enter the following:

API Key: sk-nex-your-key-here

Base URL: https://api.nexara.net/v1

For GPT / Gemini models, select or enter the standard model ID, e.g., gpt-5.4, gpt-5.2, gemini-3.1-pro.

For Claude models in Cursor, use the short model name format:

opus-4-6

sonnet-4-6

haiku-4-5

Notes

You may need to set the Override OpenAI Base URL option in Cursor

The short model name convention applies only to Claude models in Cursor

When using Claude in Cursor, do not enter the full model name with the claude- prefix or date suffix; use short names like opus-4-6, sonnet-4-6, haiku-4-5

GPT / Gemini models in Cursor use standard model IDs as usual

If the model list does not include the desired Claude entry, simply type the model name manually

Cline (VS Code Extension)

Cline is a powerful AI coding assistant for VS Code.

Installation

Open VS Code > Extensions (Ctrl+Shift+X / Cmd+Shift+X)

Search for Cline and install

After installation, the Cline icon appears in the sidebar

Configuration

Click the Cline icon > Settings (gear icon)

Set API Provider to OpenAI Compatible

Enter:

Base URL: https://api.nexara.net/v1

API Key: sk-nex-your-key-here

Model ID: claude-sonnet-4-6 (or any other standard model ID)

Claude Code

Claude Code is Anthropic's official CLI tool.

macOS / Linux

To make these persistent, add them to your shell configuration file.

Windows (Recommended: Permanent Configuration)

We recommend using Windows User Environment Variables for a one-time setup that persists across sessions.

Open the Start menu and search for Environment Variables

Select Edit environment variables for your account

Under "User variables," click New

Add the following two variables:

ANTHROPIC_BASE_URL = https://api.nexara.net

ANTHROPIC_API_KEY = sk-nex-your-key-here

Save, then close and reopen any CMD, PowerShell, or VS Code terminal windows

Verify the configuration in a new terminal window:

echo %ANTHROPIC_BASE_URL%

If you use PowerShell:

$env:ANTHROPIC_BASE_URL

If it displays https://api.nexara.net, the configuration is successful.

Launch Claude Code

If Claude Code detects a custom API Key for the first time, follow the prompt to use the key from your environment variables.

You can also configure this in ~/.claude/config.json, but the environment variable approach is more straightforward, especially for Windows users.

Important Notes

Claude Code uses the Anthropic Messages API format (not OpenAI format), so the Base URL should be https://api.nexara.net -- do not append /v1 manually.

Claude Code / the Anthropic SDK will automatically request /v1/messages as required by the protocol, so you only need to provide the root URL.

If you changed environment variables but Claude Code still reads old values, you likely need to reopen the terminal window

To switch back to the official API, delete both environment variables, reopen the terminal, and run claude login

Continue (VS Code Extension)

Install the Continue extension

Open the configuration file ~/.continue/config.json

Add AIone as a provider:

{
  "models": [
    {
      "title": "AIone Claude Sonnet",
      "provider": "openai",
      "model": "claude-sonnet-4-6",
      "apiBase": "https://api.nexara.net/v1",
      "apiKey": "sk-nex-your-key-here"
    }
  ]
}

General Configuration Principles

First determine which protocol your client uses:

Scenario	Base URL	Provider / Type
OpenAI-compatible client	`https://api.nexara.net/v1`	`OpenAI` / `OpenAI Compatible`
Anthropic native client	`https://api.nexara.net`	`Anthropic`

All scenarios use the AIone API Key (prefixed with sk-nex-).

How to Determine the Correct Model Name

If you just need to know "what model name should I enter in this client," remember these 3 rules:

For the authoritative real-time list, check GET https://api.nexara.net/v1/models and the Model List

Only Claude models in Cursor should use short model names

All other clients should use the standard model ID

For the complete rules, full model table, and Cursor Claude short name mapping, see:

Model Naming & Compatibility Rules

FAQ

Connection Failure / Timeout

For OpenAI-compatible clients, verify the Base URL is https://api.nexara.net/v1

For Claude Code / Anthropic SDK, verify the Base URL is https://api.nexara.net

Confirm your API Key is valid

Check whether your network can reach api.nexara.net

Model Not Found

Verify the model ID spelling

Confirm the model name on the Model List page

For Claude in Cursor, use short model names like opus-4-6, sonnet-4-6, haiku-4-5

For all other clients, and for GPT / Gemini, use standard model IDs like claude-sonnet-4-6, gpt-5.4, gpt-5.2, gemini-3.1-pro

For the full model name reference, see: Model Naming & Compatibility Rules

Unexpected Response Format

Ensure the Provider type is set to OpenAI Compatible (not Anthropic native format)

Exception: Claude Code uses the Anthropic format, which AIone also supports

10 - IDE Integration

IDE Integration#

Overview#

Cursor#

Setup#

Notes#

Cline (VS Code Extension)#

Installation#

Configuration#

Claude Code#

macOS / Linux#

Windows (Recommended: Permanent Configuration)#

Launch Claude Code#

Important Notes#

Continue (VS Code Extension)#

General Configuration Principles#

How to Determine the Correct Model Name#

FAQ#

Connection Failure / Timeout#

Model Not Found#

Unexpected Response Format#