Specify token override using a custom httpx client

README.md

@@ -185,7 +185,7 @@ However, Azure Container Registry has a fixed cost per registry per day.
 
 You can try the [Azure pricing calculator](https://azure.microsoft.com/pricing/calculator/) for the resources:
 
-* Azure AI Service: S0 tier, DeepSeek-R1 model. Pricing is based on token count. [Pricing]()
+* Azure AI Service: S0 tier, DeepSeek-R1 model. Pricing is based on token count. [Pricing](https://aka.ms/DeepSeekPricing)
 * Azure Container App: Consumption tier with 0.5 CPU, 1GiB memory/storage. Pricing is based on resource allocation, and each month allows for a certain amount of free usage. [Pricing](https://azure.microsoft.com/pricing/details/container-apps/)
 * Azure Container Registry: Basic tier. [Pricing](https://azure.microsoft.com/pricing/details/container-registry/)
 * Log analytics: Pay-as-you-go tier. Costs based on data ingested. [Pricing](https://azure.microsoft.com/pricing/details/monitor/)
@@ -209,5 +209,3 @@ You may want to consider additional security measures, such as:
 ### Resources
 
 * [Blog post: Building a streaming DeepSeek-R1 app on Azure](https://blog.pamelafox.org/2025/04/building-streaming-deepseek-r1-app-on.html)
-
-

src/quartapp/chat.py

@@ -1,8 +1,9 @@
 import json
 import os
 
+import httpx
 from azure.identity.aio import AzureDeveloperCliCredential, ManagedIdentityCredential, get_bearer_token_provider
-from openai import AsyncOpenAI
+from openai import AsyncOpenAI, DefaultAsyncHttpxClient
 from quart import (
     Blueprint,
     Response,
@@ -27,15 +28,25 @@ async def configure_openai():
         bp.azure_credential = AzureDeveloperCliCredential(tenant_id=tenant_id)
 
     # Get the token provider for Azure OpenAI based on the selected Azure credential
-    bp.openai_token_provider = get_bearer_token_provider(
+    openai_token_provider = get_bearer_token_provider(
         bp.azure_credential, "https://cognitiveservices.azure.com/.default"
     )
 
+    class TokenBasedAuth(httpx.Auth):
+        async def async_auth_flow(self, request):
+            token = await openai_token_provider()
+            request.headers["Authorization"] = f"Bearer {token}"
+            yield request
+
+        def sync_auth_flow(self, request):
+            raise RuntimeError("Cannot use a sync authentication class with httpx.AsyncClient")
+
     # Create the Asynchronous Azure OpenAI client
     bp.openai_client = AsyncOpenAI(
         base_url=os.environ["AZURE_INFERENCE_ENDPOINT"],
-        api_key=await bp.openai_token_provider(),
+        api_key="placeholder",
         default_query={"api-version": "2024-05-01-preview"},
+        http_client=DefaultAsyncHttpxClient(auth=TokenBasedAuth()),
     )
 
     # Set the model name to the Azure OpenAI model deployment name
@@ -63,7 +74,6 @@ async def response_stream():
             {"role": "system", "content": "You are a helpful assistant."},
         ] + request_messages
 
-        bp.openai_client.api_key = await bp.openai_token_provider()
         chat_coroutine = bp.openai_client.chat.completions.create(
             # Azure Open AI takes the deployment name as the model name
             model=bp.openai_model,