Connecting External Tools

All deployed models (local Ollama, vLLM, and remote) are accessible via an OpenAI-compatible API endpoint. Any tool that supports a custom OpenAI base URL can connect directly to your platform.

Finding Your Connection Details

Navigate to AI Models in the sidebar
Find the model you want to connect to
Click the Credentials button (key icon) on the model card
The modal shows your base URL, model ID, and API token

Connection Details

Base URL

https://<your-host>/api/v1

Replace <your-host> with your platform hostname (e.g., demo.menatic.com).

Model ID

The model ID is shown in the Credentials modal. For Ollama models it is the model name (e.g., llama3.2); for vLLM models it is the Hugging Face model ID (e.g., mistralai/Mistral-7B-Instruct-v0.3).

API Token

Your API token is a JWT issued when you log in. Retrieve it from the Credentials modal — click the copy button to copy the full token to your clipboard.

Example Configurations

curl

curl https://<your-host>/api/v1/chat/completions \
  -H "Authorization: Bearer <your-token>" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "<model-id>",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

List all available models:

curl https://<your-host>/api/v1/models \
  -H "Authorization: Bearer <your-token>"

opencode

Add the following to your ~/.config/opencode/config.json:

{
  "providers": {
    "menatic": {
      "name": "Menatic AI",
      "api": "openai",
      "baseUrl": "https://<your-host>/api/v1",
      "apiKey": "<your-token>",
      "models": {
        "<model-id>": {
          "name": "<model-name>",
          "contextLength": 32768,
          "temperature": 0.7
        }
      }
    }
  }
}

The Credentials modal provides a pre-filled config block you can copy directly.

Cursor

Open Cursor Settings → Models
Under OpenAI API Key, enter your API token
Under Override OpenAI Base URL, enter https://<your-host>/api/v1
In any chat or Composer window, type the model ID or select it from the model picker

Continue (VS Code / JetBrains)

Add a model entry to your ~/.continue/config.json:

{
  "models": [
    {
      "title": "My Model",
      "provider": "openai",
      "model": "<model-id>",
      "apiBase": "https://<your-host>/api/v1",
      "apiKey": "<your-token>"
    }
  ]
}

Supported Endpoints

Endpoint	Method	Description
`/api/v1/models`	GET	List all available models
`/api/v1/chat/completions`	POST	Chat completions (streaming supported)