Skip to main content
Version: 1.1

Platform-specific config

Cloudflare​

{
    "modelName": "@cf/qwen/qwen1.5-14b-chat-awq",
    "accountId": "<guid>",
    "apiKey": "xxx-xxx-xxx"
}

  • modelName: The name of the AI model to use, must be exist on Cloudflare and you have access to it.

  • accountId: The account ID of your Cloudflare account, it could be the tail route when you access https://dash.cloudflare.com/ .

  • apiKey: An API key issued by your account, must be valid and have permission Account.Workers AI.

OpenAI​

{
    "authCredentials": "xxx-xxx-xxx",
    "modelName": "gpt-4"
}

  • authCredentials: Authorization credentials, will be added into Autorizations header.

  • modelName: The name of the model you use.

OpenAI-compatible​

{
    "authCredentials": "xxx-xxx-xxx",
    "modelName": "gpt-4",
    "chatCompletionEndpoint": "https://.../v1/chat/completion"
}

  • authCredentials: Authorization credentials, will be added into Autorizations header.

  • modelName: The name of the model you use.

  • chatCompletionEndpoint: The URL to POST messages. If you use LM Studio, you may use http://localhost:<port>/v1/chat/completion where port is the port that LM Studio is listening on.