Azure OpenAI
Connects to Azure OpenAI to perform operations on large language models.
This article describes the operations for the Azure OpenAI built-in connector, which is available only for Standard workflows in single-tenant Azure Logic Apps.
Built-in connector settings
In a Standard logic app resource, the application and host settings control various thresholds for performance, throughput, timeout, and so on. For more information, see Edit host and app settings for Standard logic app workflows.
Connector how-to guide
For more information about integrating Azure OpenAI with your workflow in Azure Logic Apps, see Integrate Azure AI services with Standard workflows in Azure Logic Apps.
Authentication
URL and key-based authentication
Authentication type
Parameters
| Name | Description | Type | Required | Allowed Values |
|---|---|---|---|---|
| Azure OpenAI endpoint URL | The URL for the Azure OpenAI endpoint. | string | True | |
| Authentication key | The authentication key for Azure OpenAI. | securestring | True |
Active Directory OAuth
Active Directory OAuth
Parameters
| Name | Description | Type | Required | Allowed Values |
|---|---|---|---|---|
| Azure OpenAI endpoint URL | The URL for the Azure OpenAI endpoint. | string | True | |
| Active Directory OAuth | Active Directory OAuth | string | True | |
| Authority | Active Directory authority | string | False | |
| Tenant | Active Directory tenant | string | True | |
| Credential type | Active Directory credential type | string | False | Certificate, Secret |
| Client ID | Active Directory client ID | string | True | |
| Client secret | Active Directory client secret | securestring | True | |
| Pfx | Active Directory pfx | securestring | True | |
| Password | Active Directory password | securestring | True |
Managed identity
Managed identity
Parameters
| Name | Description | Type | Required | Allowed Values |
|---|---|---|---|---|
| Azure OpenAI endpoint URL | The URL for the Azure OpenAI endpoint. | string | True | |
| Managed identity | Managed identity | string | True | |
| Managed identity | Managed identity | string | False |
Actions
| Get an embedding |
Executes a GetEmbeddings request for the specified single text input. |
| Get chat completions |
Gets chat completions for the specified text input. |
| Get completion (Preview) |
Gets completion for the specified text prompt. |
| Get multiple chat completions (Preview) |
Get multiple chat completion options for the specified text input. |
| Get multiple embeddings |
Executes a GetEmbeddings request for the specified array of text inputs. |
Get an embedding
Executes a GetEmbeddings request for the specified single text input.
Parameters
| Name | Key | Required | Type | Description |
|---|---|---|---|---|
|
Deployment identifier
|
deploymentId | True | string |
The deployment or model name. |
|
Single text input
|
input | True | string |
The single text to convert to an embedding. |
Returns
| Name | Path | Type | Description |
|---|---|---|---|
|
Embedding
|
embedding | string |
An array of floats that represent the input's computed embeddings. |
Get chat completions
Gets chat completions for the specified text input.
Parameters
| Name | Key | Required | Type | Description |
|---|---|---|---|---|
|
Deployment identifier
|
deploymentId | True | string |
The deployment or model name. |
|
Sampling temperature
|
temperature | string |
A value used to control the apparent creativity of generated completions. See Azure.AI.Inference.ChatCompletionsOptions.Temperature Property. |
|
|
messages
|
messages | True | string |
The deployment or model name. |
|
Nucleus sampling (top_p)
|
top_p | string |
A value used to control the apparent creativity of generated completions and an alternative value to Temperature. See ChatCompletionsOptions.NucleusSamplingFactor Property. |
|
|
Max tokens
|
max_tokens | string |
The maximum number of tokens to generate. See ChatCompletionsOptions.MaxTokens Property. |
|
|
Presence penalty
|
presence_penalty | string |
A value that influences the probability of generated tokens appearing based on their existing presence in generated text. See ChatCompletionsOptions.PresencePenalty Property. |
|
|
Frequency penalty
|
frequency_penalty | string |
A value that influences the probability of generated tokens appearing based on their cumulative frequence. See ChatCompletionsOptions.FrequencyPenalty Property. |
Returns
| Name | Path | Type | Description |
|---|---|---|---|
|
Chat completion output role
|
role | string |
The chat completion output role. |
|
Chat completion response
|
content | string |
The chat completion response. |
|
Chat user
|
name | string |
The chat user. |
Get completion (Preview)
Gets completion for the specified text prompt.
Parameters
| Name | Key | Required | Type | Description |
|---|---|---|---|---|
|
Deployment identifier
|
deploymentId | True | string |
The deployment or model name. |
|
Sampling temperature
|
temperature | string |
A value used to control the apparent creativity of generated completions. See Azure.AI.Inference.ChatCompletionsOptions.Temperature Property. |
|
|
prompts
|
prompts | True | string |
The deployment or model name. |
|
stopSequences
|
stopSequences | string |
The deployment or model name. |
|
|
Max tokens
|
max_tokens | string |
The maximum number of tokens to generate. See ChatCompletionsOptions.MaxTokens Property. |
|
|
Presence penalty
|
presence_penalty | string |
A value that influences the probability of generated tokens appearing based on their existing presence in generated text. See ChatCompletionsOptions.PresencePenalty Property. |
|
|
Frequency penalty
|
frequency_penalty | string |
A value that influences the probability of generated tokens appearing based on their cumulative frequence. See ChatCompletionsOptions.FrequencyPenalty Property. |
Returns
| Name | Path | Type | Description |
|---|---|---|---|
|
Completion response
|
text | string |
The completion response. |
Get multiple chat completions (Preview)
Get multiple chat completion options for the specified text input.
Parameters
| Name | Key | Required | Type | Description |
|---|---|---|---|---|
|
Deployment identifier
|
deploymentId | True | string |
The deployment or model name. |
|
Sampling temperature
|
temperature | string |
A value used to control the apparent creativity of generated completions. See Azure.AI.Inference.ChatCompletionsOptions.Temperature Property. |
|
|
messages
|
messages | True | string |
The deployment or model name. |
|
Nucleus sampling (top_p)
|
top_p | string |
A value used to control the apparent creativity of generated completions and an alternative value to Temperature. See ChatCompletionsOptions.NucleusSamplingFactor Property. |
|
|
Max tokens
|
max_tokens | string |
The maximum number of tokens to generate. See ChatCompletionsOptions.MaxTokens Property. |
|
|
Choice count (n)
|
n | string |
The number of chat responses to generate. See ChatCompletions.Choices Property. |
|
|
Presence penalty
|
presence_penalty | string |
A value that influences the probability of generated tokens appearing based on their existing presence in generated text. See ChatCompletionsOptions.PresencePenalty Property. |
|
|
Frequency penalty
|
frequency_penalty | string |
A value that influences the probability of generated tokens appearing based on their cumulative frequence. See ChatCompletionsOptions.FrequencyPenalty Property. |
Returns
| Name | Path | Type | Description |
|---|---|---|---|
|
Chat message choices
|
choices | string |
The array of chat message choices. |
Get multiple embeddings
Executes a GetEmbeddings request for the specified array of text inputs.
Parameters
| Name | Key | Required | Type | Description |
|---|---|---|---|---|
|
Deployment identifier
|
deploymentId | True | string |
The deployment or model name. |
|
Array input
|
input | True | string |
The array input to convert to embeddings. |
Returns
| Name | Path | Type | Description |
|---|---|---|---|
|
Embeddings array
|
embeddings | string |
An array with arrays of floats representing the input's computed embeddings. |