Struct AzureOpenAIProvider 

pub struct AzureOpenAIProvider { /* private fields */ }

Azure OpenAI provider implementation.

This provider creates language models, embedding models, and image models that use Azure OpenAI endpoints. It handles Azure-specific authentication (api-key header) and URL formatting (deployment-based or v1 API).

Examples

Using Builder Pattern (Recommended)

use llm_kit_azure::AzureClient;

let provider = AzureClient::new()
    .resource_name("my-azure-resource")
    .api_key("your-api-key")
    .build();

let model = provider.chat_model("gpt-4-deployment");

Direct Instantiation (Alternative)

use llm_kit_azure::{AzureOpenAIProvider, AzureOpenAIProviderSettings};

let provider = AzureOpenAIProvider::new(
    AzureOpenAIProviderSettings::new()
        .with_resource_name("my-azure-resource")
        .with_api_key("your-api-key")
);

let model = provider.chat_model("gpt-4-deployment");

Implementations

impl AzureOpenAIProvider

pub fn new(settings: AzureOpenAIProviderSettings) -> Self

Creates a new Azure OpenAI provider.

Panics

Panics if the settings are invalid (missing both base_url and resource_name).

pub fn model(&self, deployment_id: impl Into<String>) -> Arc<dyn LanguageModel>

Creates a language model with the given deployment ID.

This is an alias for chat_model().

pub fn chat_model( &self, deployment_id: impl Into<String>, ) -> Arc<dyn LanguageModel>

Creates a chat language model with the given deployment ID.

Arguments

• deployment_id - The deployment name/ID in Azure OpenAI

pub fn completion_model( &self, deployment_id: impl Into<String>, ) -> Arc<dyn LanguageModel>

Creates a completion language model with the given deployment ID.

Arguments

• deployment_id - The deployment name/ID in Azure OpenAI

pub fn text_embedding_model( &self, deployment_id: impl Into<String>, ) -> Arc<dyn EmbeddingModel<String>>

Creates a text embedding model with the given deployment ID.

Arguments

• deployment_id - The deployment name/ID in Azure OpenAI

pub fn image_model( &self, deployment_id: impl Into<String>, ) -> Arc<dyn ImageModel>

Creates an image model with the given deployment ID.

Arguments

• deployment_id - The deployment name/ID in Azure OpenAI (e.g., dall-e-3 deployment)

pub fn name(&self) -> &str

Gets the provider name.

Trait Implementations

impl Provider for AzureOpenAIProvider

fn language_model( &self, deployment_id: &str, ) -> Result<Arc<dyn LanguageModel>, ProviderError>

Returns the language model with the given id.

fn text_embedding_model( &self, deployment_id: &str, ) -> Result<Arc<dyn EmbeddingModel<String>>, ProviderError>

Returns the text embedding model with the given id.

fn image_model( &self, deployment_id: &str, ) -> Result<Arc<dyn ImageModel>, ProviderError>

Returns the image model with the given id.

fn transcription_model( &self, deployment_id: &str, ) -> Result<Arc<dyn TranscriptionModel>, ProviderError>

Returns the transcription model with the given id.

fn speech_model( &self, deployment_id: &str, ) -> Result<Arc<dyn SpeechModel>, ProviderError>

Returns the speech model with the given id.

fn reranking_model( &self, deployment_id: &str, ) -> Result<Arc<dyn RerankingModel>, ProviderError>

Returns the reranking model with the given id.

fn specification_version(&self) -> &str

The provider must specify which provider interface version it implements. This will allow us to evolve the provider interface and retain backwards compatibility.