Hi I would like to know apart from setting the open ai ali key in integration, what are the next steps for assistant to do model calls via our own open ai api

After setting your OpenAI API key in the integration, the next step is to configure your assistant to use your own OpenAI endpoint. Here’s a sample JSON configuration:

json
{
  "name": "My Assistant",
  "model": {
    "provider": "custom-llm",
    "url": "https://api.openai.com/v1",
    "model": "your-model-name",
    "messages": [
      { "role": "system", "content": "You are an assistant." }
    ],
    "temperature": 0.7
  }
}

This will route model calls through your own OpenAI API key and endpoint. For more details, see the [Vapi documentation](https://docs.vapi.ai/customization/custom-llm/fine-tuned-openai-models). Source: - [Fine-tuned OpenAI models](https://docs.vapi.ai/customization/custom-llm/fine-tuned-openai-models)

How to setup this in vapi dashboard?

To set up your own OpenAI API routing in the Vapi dashboard: 1. Go to the Vapi Dashboard and open your assistant. 2. In the Model section, select Custom LLM as the provider. 3. Paste your server's public URL (e.g., from Ngrok) in the endpoint field. 4. Click Test Connection to verify setup. This will route model calls through your own OpenAI API server. For more details, see the [official guide](https://docs.vapi.ai/customization/custom-llm/using-your-server). Source: - [Using your server for custom LLMs](https://docs.vapi.ai/customization/custom-llm/using-your-server)

@Vapi Support Bot any option to setup open api key in vapi dashboard and route calls without running own open ai model server

Yes, you can use your OpenAI API key directly in the Vapi dashboard without running your own server. Just set up your assistant's model like this in the dashboard:

json
{
  "provider": "custom-llm",
  "url": "https://api.openai.com/v1",
  "model": "your-model-name"
}

Then, assign your OpenAI API key in the integration settings. This will route all model calls through your OpenAI account automatically—no server needed. For more details, see the official guide ([Fine-tuned OpenAI models](https://docs.vapi.ai/customization/custom-llm/fine-tuned-openai-models)). Source: - [Fine-tuned OpenAI models](https://docs.vapi.ai/customization/custom-llm/fine-tuned-openai-models)