Skip to main content

​
What are AI Application Integrations?

Instead of writing code yourself, you can connect Model Router to existing AI development tools. These tools provide user-friendly interfaces for AI-assisted coding and development.
Before configuring these tools, make sure you have:
  1. Created an API key (see Create API Key)
  2. Checked which models are available (see List Available Models)

​
Supported Applications

​
Claude Code

Claude Code is a code editor integration tool based on Anthropic Claude. It provides AI-powered coding assistance directly in your terminal.

​
Codex CLI

Codex CLI is a command-line interface AI code assistant tool that helps you write and understand code.

​
Cherry Studio

Cherry Studio is a powerful desktop AI client with multi-model conversation and drawing capabilities. It integrates 30+ industry-specific intelligent assistants.

​
Claude Code Integration

​
What You Need

  • ANTHROPIC_AUTH_TOKEN: Your API key from Memorylake Console (starts with sk-)
  • ANTHROPIC_BASE_URL: Model Router service address: https://app.memorylake.ai/claude

​
Step 1: Install Claude Code

Install Claude Code globally via npm: 
npm install -g @anthropic-ai/claude-code
Verify installation: 
claude --version

​
Step 2: Get Your API Key

Click “Add API Key” on the API key management page in the Memorylake Console to get an API Key.
The name can be set arbitrarily. It’s recommended to set the quota to unlimited. Keep other settings as default.
Get API Key from Memorylake Console

​
Step 3: Configure Environment Variables

Set environment variables in your project directory: 
cd your-project-folder
export ANTHROPIC_AUTH_TOKEN=sk-...
export ANTHROPIC_BASE_URL=https://app.memorylake.ai/claude
claude

​
Step 4: Initialize Configuration

After running, complete initialization as prompted:
  1. Select your preferred theme + Enter
  2. Confirm safety notice + Enter
  3. Use default Terminal configuration + Enter
  4. Trust working directory + Enter

​
Step 5: Start Using

After configuration, you can code with your AI programming partner in the terminal.
Claude Code usage interface
To avoid repeatedly entering environment variables each time, you can write the configuration to shell configuration files: Bash Configuration: 
echo -e '\n export ANTHROPIC_AUTH_TOKEN=sk-...' >> ~/.bash_profile
echo -e '\n export ANTHROPIC_BASE_URL=https://app.memorylake.ai/claude' >> ~/.bash_profile
echo -e '\n export ANTHROPIC_AUTH_TOKEN=sk-...' >> ~/.bashrc
echo -e '\n export ANTHROPIC_BASE_URL=https://app.memorylake.ai/claude' >> ~/.bashrc
Zsh Configuration: 
echo -e '\n export ANTHROPIC_AUTH_TOKEN=sk-...' >> ~/.zshrc
echo -e '\n export ANTHROPIC_BASE_URL=https://app.memorylake.ai/claude' >> ~/.zshrc
After configuration, restart the terminal to use directly: 
cd your-project-folder
claude

​
Codex CLI Integration

​
What You Need

  • OPENAI_API_KEY: Your API key from Memorylake Console (starts with sk-)
  • base_url: Model Router service address: https://app.memorylake.ai/v1
  • model: Model name to use (check available models)

​
Step 1: Install Node.js

Ensure Node.js environment is installed. Codex CLI requires Node.js support.

​
Step 2: Install Codex CLI

Install Codex CLI globally via npm: 
npm i -g @openai/codex
Verify installation: 
codex --version

​
Step 3: Get Your API Key

Click “Add API Key” on the API key management page in the Memorylake Console to get an API Key.
The name can be set arbitrarily. It’s recommended to set the quota to unlimited. Keep other settings as default.
Get API Key from Memorylake Console

​
Step 4: Create Configuration Files

Create ~/.codex/config.toml file and add the following configuration:
config.toml
model = "gpt-5"
model_provider = "memorylake"
preferred_auth_method = "apikey"

[model_providers.memorylake]
name = "MemoryLake"
base_url = "https://app.memorylake.ai/v1"
wire_api = "responses"
Create ~/.codex/auth.json file and add the following configuration:
auth.json
{
  "OPENAI_API_KEY": "Replace here with your KEY"
}
The configuration file path ~/.codex can also be specified using the CODEX_HOME environment variable.

​
Step 5: Start Using

Run in the project directory: 
cd your-project-folder
codex
Codex CLI usage interface

​
Configuration File Details

​
Codex CLI config.toml

  • model: Specify the model name to use (must be available to your API key)
  • model_provider: Model provider identifier (use memorylake to represent Model Router)
  • preferred_auth_method: Authentication method (use apikey)
  • base_url: Model Router API address, must include /v1 path

​
Codex CLI auth.json

​
Custom Configuration Path

To use a custom configuration path, set the CODEX_HOME environment variable: 
export CODEX_HOME=/path/to/your/config

​
Cherry Studio Integration

​
What You Need

  • Provider Type: Select a type supported by Model Router
  • API Key: Your API key from Memorylake Console (starts with sk-)
  • API Address: Model Router service address

​
Step 1: Get Your API Key

Copy your API Key from the API key management page in the Memorylake Console.
Copy API Key from Memorylake Console

​
Step 2: Add Provider in Cherry Studio

  1. Open Cherry Studio
  2. Add a new Provider
  3. Select the corresponding Provider Type
  4. Enter your API Key and API Address
In the chat settings under system settings in the Memorylake Console, you can add a quick option to one-click fill Cherry Studio configuration on the API key management page.

​
Step 3: Add Models

After configuration, add available models to Cherry Studio. Make sure to check List Available Models to see which models you can use.

​
Step 4: Start Using

Return to the chat page and select the configured Model Router model to start a conversation. You can switch between different Model Router models at any time.

​
Common Issues

​
Cherry Studio

  • Provider Not Working: Check that the API address and API Key are correct
  • Models Not Available: Confirm that the models are enabled in your Model Router API key group. Check List Available Models
  • Connection Failed: Verify your network connection and API endpoint URL

​
Claude Code

  • Installation Failed: Ensure Node.js and npm are installed; using npm install -g requires administrator privileges.
  • Connection Failed: Check if the API address and API Key are correct, and ensure network connection is normal.
  • Environment Variables Not Effective: Confirm they are correctly written to the configuration file and restart the terminal or execute source ~/.bashrc (or source ~/.zshrc).

​
Codex CLI

  • Installation Failed: Ensure Node.js and npm are installed; using npm i -g requires administrator privileges.
  • Configuration File Not Found: Manually create ~/.codex directory and configuration files.
  • Connection Failed: Check if base_url and OPENAI_API_KEY are correct, and ensure network connection is normal.
  • Model Unavailable: Confirm that the model is enabled and configured in your Model Router API key group. Check List Available Models.

​
Important Notes

  1. Model Availability: Model availability depends on your Model Router API key group configuration. Always check List Available Models first.
  2. API Key Security: Never share your API key publicly or commit it to version control. Use environment variables or secure configuration files.
  3. Quota Management: Monitor your usage in the console. See View Usage and Billing for details.