Custom headers

(Optional) Static headers sent with every request. Variable expansion is supported.

Deployments

Deployment name

Required

Deployment type

Model handling mode

API Mode

Choose whether the model will use Chat Completions API or Responses API.

Max tokens API mode

Model pricing

Use the pricing from a given OpenAI model, or enter a custom pricing

Prompt cost

In $/1k tokens

Completion cost

In $/1k tokens

Embedding cost

In $/1k tokens

Fine-tuning

Allow fine-tuning
Enable to allow models from this connection in the fine-tuning recipe.

Resource group

Subscription Id

AzureML connection
Its credentials are used to deploy fine-tuned models with Azure Management API

You need to input a resource group and a subscription id if you want to be able to deploy your Azure Open AI models.

Fine-tunable models

GPT 3.5 Chat turbo (0125)

GPT 3.5 Chat turbo (0613)

GPT 3.5 Chat turbo (1106)

Babbage 002

Davinci 002

Custom Model Id

Required

Custom model display name

Required

Use chat API

Can be fine-tuned

Credentials mode