openaivec 0.13.0__tar.gz → 0.13.2__tar.gz

{openaivec-0.13.0 → openaivec-0.13.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: openaivec
-Version: 0.13.0
+Version: 0.13.2
 Summary: Generative mutation for tabular calculation
 Project-URL: Homepage, https://microsoft.github.io/openaivec/
 Project-URL: Repository, https://github.com/microsoft/openaivec
@@ -180,8 +180,8 @@ from openaivec import pandas_ext
 os.environ["OPENAI_API_KEY"] = "your-api-key-here"
 # Or for Azure OpenAI:
 # os.environ["AZURE_OPENAI_API_KEY"] = "your-azure-key"
-# os.environ["AZURE_OPENAI_API_ENDPOINT"] = "https://<your-resource-name>.services.ai.azure.com"
-# os.environ["AZURE_OPENAI_API_VERSION"] = "2025-04-01-preview"
+# os.environ["AZURE_OPENAI_BASE_URL"] = "https://YOUR-RESOURCE-NAME.services.ai.azure.com/openai/v1/"
+# os.environ["AZURE_OPENAI_API_VERSION"] = "preview"
 
 # Authentication Option 2: Custom client (optional)
 # from openai import OpenAI, AsyncOpenAI
@@ -190,6 +190,7 @@ os.environ["OPENAI_API_KEY"] = "your-api-key-here"
 # pandas_ext.use_async(AsyncOpenAI())
 
 # Configure model (optional - defaults to gpt-4.1-mini)
+# For Azure OpenAI: use your deployment name, for OpenAI: use model name
 pandas_ext.responses_model("gpt-4.1-mini")
 
 # Create your data
@@ -211,6 +212,27 @@ result = df.assign(
 
 📓 **[Interactive pandas examples →](https://microsoft.github.io/openaivec/examples/pandas/)**
 
+### Using with Reasoning Models
+
+When using reasoning models (o1-preview, o1-mini, o3-mini, etc.), you must set `temperature=None` to avoid API errors:
+
+```python
+# For reasoning models like o1-preview, o1-mini, o3-mini
+pandas_ext.responses_model("o1-mini")  # Set your reasoning model
+
+# MUST use temperature=None with reasoning models
+result = df.assign(
+    analysis=lambda df: df.text.ai.responses(
+        "Analyze this text step by step",
+        temperature=None  # Required for reasoning models
+    )
+)
+```
+
+**Why this is needed**: Reasoning models don't support temperature parameters and will return an error if temperature is specified. The library automatically detects these errors and provides guidance on how to fix them.
+
+**Reference**: [Azure OpenAI Reasoning Models](https://learn.microsoft.com/en-us/azure/ai-foundry/openai/how-to/reasoning)
+
 ### Using Pre-configured Tasks
 
 For common text processing operations, openaivec provides ready-to-use tasks that eliminate the need to write custom prompts:
@@ -322,7 +344,7 @@ sc.environment["OPENAI_API_KEY"] = os.environ.get("OPENAI_API_KEY")
 
 # Option 2: Using Azure OpenAI
 # sc.environment["AZURE_OPENAI_API_KEY"] = os.environ.get("AZURE_OPENAI_API_KEY")
-# sc.environment["AZURE_OPENAI_API_ENDPOINT"] = os.environ.get("AZURE_OPENAI_API_ENDPOINT")
+# sc.environment["AZURE_OPENAI_BASE_URL"] = os.environ.get("AZURE_OPENAI_BASE_URL")
 # sc.environment["AZURE_OPENAI_API_VERSION"] = os.environ.get("AZURE_OPENAI_API_VERSION")
 ```
 
@@ -380,6 +402,16 @@ spark.udf.register(
     )
 )
 
+# --- Register UDF for Reasoning Models ---
+# For reasoning models (o1-preview, o1-mini, o3, etc.), set temperature=None
+spark.udf.register(
+    "reasoning_analysis",
+    responses_udf(
+        instructions="Analyze this step by step with detailed reasoning",
+        temperature=None  # Required for reasoning models
+    )
+)
+
 ```
 
 You can now use these UDFs in Spark SQL:
@@ -666,15 +698,15 @@ steps:
 
      # Configure Azure OpenAI authentication
      sc.environment["AZURE_OPENAI_API_KEY"] = "<your-api-key>"
-     sc.environment["AZURE_OPENAI_API_ENDPOINT"] = "https://<your-resource-name>.services.ai.azure.com"
-     sc.environment["AZURE_OPENAI_API_VERSION"] = "2025-04-01-preview"
+     sc.environment["AZURE_OPENAI_BASE_URL"] = "https://YOUR-RESOURCE-NAME.services.ai.azure.com/openai/v1/"
+     sc.environment["AZURE_OPENAI_API_VERSION"] = "preview"
 
      # Register UDFs
      spark.udf.register(
          "analyze_text",
          responses_udf(
              instructions="Analyze the sentiment of the text",
-             model_name="<your-deployment-name>"
+             model_name="gpt-4.1-mini"  # Use your Azure deployment name here
          )
      )
      ```

{openaivec-0.13.0 → openaivec-0.13.2}/README.md

@@ -156,8 +156,8 @@ from openaivec import pandas_ext
 os.environ["OPENAI_API_KEY"] = "your-api-key-here"
 # Or for Azure OpenAI:
 # os.environ["AZURE_OPENAI_API_KEY"] = "your-azure-key"
-# os.environ["AZURE_OPENAI_API_ENDPOINT"] = "https://<your-resource-name>.services.ai.azure.com"
-# os.environ["AZURE_OPENAI_API_VERSION"] = "2025-04-01-preview"
+# os.environ["AZURE_OPENAI_BASE_URL"] = "https://YOUR-RESOURCE-NAME.services.ai.azure.com/openai/v1/"
+# os.environ["AZURE_OPENAI_API_VERSION"] = "preview"
 
 # Authentication Option 2: Custom client (optional)
 # from openai import OpenAI, AsyncOpenAI
@@ -166,6 +166,7 @@ os.environ["OPENAI_API_KEY"] = "your-api-key-here"
 # pandas_ext.use_async(AsyncOpenAI())
 
 # Configure model (optional - defaults to gpt-4.1-mini)
+# For Azure OpenAI: use your deployment name, for OpenAI: use model name
 pandas_ext.responses_model("gpt-4.1-mini")
 
 # Create your data
@@ -187,6 +188,27 @@ result = df.assign(
 
 📓 **[Interactive pandas examples →](https://microsoft.github.io/openaivec/examples/pandas/)**
 
+### Using with Reasoning Models
+
+When using reasoning models (o1-preview, o1-mini, o3-mini, etc.), you must set `temperature=None` to avoid API errors:
+
+```python
+# For reasoning models like o1-preview, o1-mini, o3-mini
+pandas_ext.responses_model("o1-mini")  # Set your reasoning model
+
+# MUST use temperature=None with reasoning models
+result = df.assign(
+    analysis=lambda df: df.text.ai.responses(
+        "Analyze this text step by step",
+        temperature=None  # Required for reasoning models
+    )
+)
+```
+
+**Why this is needed**: Reasoning models don't support temperature parameters and will return an error if temperature is specified. The library automatically detects these errors and provides guidance on how to fix them.
+
+**Reference**: [Azure OpenAI Reasoning Models](https://learn.microsoft.com/en-us/azure/ai-foundry/openai/how-to/reasoning)
+
 ### Using Pre-configured Tasks
 
 For common text processing operations, openaivec provides ready-to-use tasks that eliminate the need to write custom prompts:
@@ -298,7 +320,7 @@ sc.environment["OPENAI_API_KEY"] = os.environ.get("OPENAI_API_KEY")
 
 # Option 2: Using Azure OpenAI
 # sc.environment["AZURE_OPENAI_API_KEY"] = os.environ.get("AZURE_OPENAI_API_KEY")
-# sc.environment["AZURE_OPENAI_API_ENDPOINT"] = os.environ.get("AZURE_OPENAI_API_ENDPOINT")
+# sc.environment["AZURE_OPENAI_BASE_URL"] = os.environ.get("AZURE_OPENAI_BASE_URL")
 # sc.environment["AZURE_OPENAI_API_VERSION"] = os.environ.get("AZURE_OPENAI_API_VERSION")
 ```
 
@@ -356,6 +378,16 @@ spark.udf.register(
     )
 )
 
+# --- Register UDF for Reasoning Models ---
+# For reasoning models (o1-preview, o1-mini, o3, etc.), set temperature=None
+spark.udf.register(
+    "reasoning_analysis",
+    responses_udf(
+        instructions="Analyze this step by step with detailed reasoning",
+        temperature=None  # Required for reasoning models
+    )
+)
+
 ```
 
 You can now use these UDFs in Spark SQL:
@@ -642,15 +674,15 @@ steps:
 
      # Configure Azure OpenAI authentication
      sc.environment["AZURE_OPENAI_API_KEY"] = "<your-api-key>"
-     sc.environment["AZURE_OPENAI_API_ENDPOINT"] = "https://<your-resource-name>.services.ai.azure.com"
-     sc.environment["AZURE_OPENAI_API_VERSION"] = "2025-04-01-preview"
+     sc.environment["AZURE_OPENAI_BASE_URL"] = "https://YOUR-RESOURCE-NAME.services.ai.azure.com/openai/v1/"
+     sc.environment["AZURE_OPENAI_API_VERSION"] = "preview"
 
      # Register UDFs
      spark.udf.register(
          "analyze_text",
          responses_udf(
              instructions="Analyze the sentiment of the text",
-             model_name="<your-deployment-name>"
+             model_name="gpt-4.1-mini"  # Use your Azure deployment name here
          )
      )
      ```

{openaivec-0.13.0 → openaivec-0.13.2}/src/openaivec/embeddings.py

@@ -4,7 +4,7 @@ from typing import List
 
 import numpy as np
 from numpy.typing import NDArray
-from openai import AsyncOpenAI, OpenAI, RateLimitError
+from openai import AsyncOpenAI, InternalServerError, OpenAI, RateLimitError
 
 from .log import observe
 from .proxy import AsyncBatchingMapProxy, BatchingMapProxy
@@ -24,7 +24,7 @@ class BatchEmbeddings:
 
     Attributes:
         client (OpenAI): Configured OpenAI client.
-        model_name (str): Model identifier (e.g., ``"text-embedding-3-small"``).
+        model_name (str): For Azure OpenAI, use your deployment name. For OpenAI, use the model name (e.g., ``"text-embedding-3-small"``).
         cache (BatchingMapProxy[str, NDArray[np.float32]]): Batching proxy for ordered, cached mapping.
     """
 
@@ -38,7 +38,7 @@ class BatchEmbeddings:
 
         Args:
             client (OpenAI): OpenAI client.
-            model_name (str): Embeddings model name.
+            model_name (str): For Azure OpenAI, use your deployment name. For OpenAI, use the model name.
             batch_size (int, optional): Max unique inputs per API call. Defaults to 128.
 
         Returns:
@@ -47,7 +47,7 @@ class BatchEmbeddings:
         return cls(client=client, model_name=model_name, cache=BatchingMapProxy(batch_size=batch_size))
 
     @observe(_LOGGER)
-    @backoff(exception=RateLimitError, scale=15, max_retries=8)
+    @backoff(exceptions=[RateLimitError, InternalServerError], scale=1, max_retries=12)
     def _embed_chunk(self, inputs: List[str]) -> List[NDArray[np.float32]]:
         """Embed one minibatch of strings.
 
@@ -90,7 +90,7 @@ class AsyncBatchEmbeddings:
         import asyncio
         import numpy as np
         from openai import AsyncOpenAI
-    from openaivec import AsyncBatchEmbeddings
+        from openaivec import AsyncBatchEmbeddings
 
         # Assuming openai_async_client is an initialized AsyncOpenAI client
         openai_async_client = AsyncOpenAI() # Replace with your actual client initialization
@@ -119,7 +119,7 @@ class AsyncBatchEmbeddings:
 
     Attributes:
         client (AsyncOpenAI): Configured OpenAI async client.
-        model_name (str): Embeddings model name.
+        model_name (str): For Azure OpenAI, use your deployment name. For OpenAI, use the model name.
         cache (AsyncBatchingMapProxy[str, NDArray[np.float32]]): Async batching proxy.
     """
 
@@ -141,7 +141,7 @@ class AsyncBatchEmbeddings:
 
         Args:
             client (AsyncOpenAI): OpenAI async client.
-            model_name (str): Embeddings model name.
+            model_name (str): For Azure OpenAI, use your deployment name. For OpenAI, use the model name.
             batch_size (int, optional): Max unique inputs per API call. Defaults to 128.
             max_concurrency (int, optional): Max concurrent API calls. Defaults to 8.
 
@@ -155,7 +155,7 @@ class AsyncBatchEmbeddings:
         )
 
     @observe(_LOGGER)
-    @backoff_async(exception=RateLimitError, scale=15, max_retries=8)
+    @backoff_async(exceptions=[RateLimitError, InternalServerError], scale=1, max_retries=12)
     async def _embed_chunk(self, inputs: List[str]) -> List[NDArray[np.float32]]:
         """Embed one minibatch of strings asynchronously.
 

{openaivec-0.13.0 → openaivec-0.13.2}/src/openaivec/model.py

@@ -59,29 +59,65 @@ class PreparedTask:
 
 @dataclass(frozen=True)
 class ResponsesModelName:
+    """Container for responses model name configuration.
+
+    Attributes:
+        value (str): The model name for OpenAI responses API.
+    """
+
     value: str
 
 
 @dataclass(frozen=True)
 class EmbeddingsModelName:
+    """Container for embeddings model name configuration.
+
+    Attributes:
+        value (str): The model name for OpenAI embeddings API.
+    """
+
     value: str
 
 
 @dataclass(frozen=True)
 class OpenAIAPIKey:
+    """Container for OpenAI API key configuration.
+
+    Attributes:
+        value (str): The API key for OpenAI services.
+    """
+
     value: str
 
 
 @dataclass(frozen=True)
 class AzureOpenAIAPIKey:
+    """Container for Azure OpenAI API key configuration.
+
+    Attributes:
+        value (str): The API key for Azure OpenAI services.
+    """
+
     value: str
 
 
 @dataclass(frozen=True)
-class AzureOpenAIEndpoint:
+class AzureOpenAIBaseURL:
+    """Container for Azure OpenAI base URL configuration.
+
+    Attributes:
+        value (str): The base URL for Azure OpenAI services.
+    """
+
     value: str
 
 
 @dataclass(frozen=True)
 class AzureOpenAIAPIVersion:
+    """Container for Azure OpenAI API version configuration.
+
+    Attributes:
+        value (str): The API version for Azure OpenAI services.
+    """
+
     value: str

{openaivec-0.13.0 → openaivec-0.13.2}/src/openaivec/pandas_ext.py

@@ -7,7 +7,7 @@ from openaivec import pandas_ext
 
 # Option 1: Use environment variables (automatic detection)
 # Set OPENAI_API_KEY or Azure OpenAI environment variables
-# (AZURE_OPENAI_API_KEY, AZURE_OPENAI_API_ENDPOINT, AZURE_OPENAI_API_VERSION)
+# (AZURE_OPENAI_API_KEY, AZURE_OPENAI_BASE_URL, AZURE_OPENAI_API_VERSION)
 # No explicit setup needed - clients are automatically created
 
 # Option 2: Use an existing OpenAI client instance
@@ -17,14 +17,18 @@ pandas_ext.use(client)
 # Option 3: Use an existing Azure OpenAI client instance
 azure_client = AzureOpenAI(
     api_key="your-azure-key",
-    azure_endpoint="https://<your-resource-name>.services.ai.azure.com",
-    api_version="2025-04-01-preview"
+    base_url="https://YOUR-RESOURCE-NAME.services.ai.azure.com/openai/v1/",
+    api_version="preview"
 )
 pandas_ext.use(azure_client)
 
-# Option 4: Use async clients
-async_client = AsyncOpenAI(api_key="your-api-key")
-pandas_ext.use_async(async_client)
+# Option 4: Use async Azure OpenAI client instance
+async_azure_client = AsyncAzureOpenAI(
+    api_key="your-azure-key",
+    base_url="https://YOUR-RESOURCE-NAME.services.ai.azure.com/openai/v1/",
+    api_version="preview"
+)
+pandas_ext.use_async(async_azure_client)
 
 # Set up model names (optional, defaults shown)
 pandas_ext.responses_model("gpt-4.1-mini")
@@ -48,7 +52,7 @@ from pydantic import BaseModel
 
 from .embeddings import AsyncBatchEmbeddings, BatchEmbeddings
 from .model import EmbeddingsModelName, PreparedTask, ResponseFormat, ResponsesModelName
-from .provider import CONTAINER
+from .provider import CONTAINER, _check_azure_v1_api_url
 from .proxy import AsyncBatchingMapProxy, BatchingMapProxy
 from .responses import AsyncBatchResponses, BatchResponses
 from .task.table import FillNaResponse, fillna
@@ -74,6 +78,10 @@ def use(client: OpenAI) -> None:
             `openai.AzureOpenAI` instance.
             The same instance is reused by every helper in this module.
     """
+    # Check Azure v1 API URL if using AzureOpenAI client
+    if client.__class__.__name__ == "AzureOpenAI" and hasattr(client, "base_url"):
+        _check_azure_v1_api_url(str(client.base_url))
+
     CONTAINER.register(OpenAI, lambda: client)
 
 
@@ -85,6 +93,10 @@ def use_async(client: AsyncOpenAI) -> None:
             `openai.AsyncAzureOpenAI` instance.
             The same instance is reused by every helper in this module.
     """
+    # Check Azure v1 API URL if using AsyncAzureOpenAI client
+    if client.__class__.__name__ == "AsyncAzureOpenAI" and hasattr(client, "base_url"):
+        _check_azure_v1_api_url(str(client.base_url))
+
     CONTAINER.register(AsyncOpenAI, lambda: client)
 
 
@@ -92,7 +104,7 @@ def responses_model(name: str) -> None:
     """Override the model used for text responses.
 
     Args:
-        name (str): Model name as listed in the OpenAI API
+        name (str): For Azure OpenAI, use your deployment name. For OpenAI, use the model name
             (for example, ``gpt-4.1-mini``).
     """
     CONTAINER.register(ResponsesModelName, lambda: ResponsesModelName(name))
@@ -102,7 +114,8 @@ def embeddings_model(name: str) -> None:
     """Override the model used for text embeddings.
 
     Args:
-        name (str): Embedding model name, e.g. ``text-embedding-3-small``.
+        name (str): For Azure OpenAI, use your deployment name. For OpenAI, use the model name,
+            e.g. ``text-embedding-3-small``.
     """
     CONTAINER.register(EmbeddingsModelName, lambda: EmbeddingsModelName(name))
 
@@ -143,7 +156,7 @@ class OpenAIVecSeriesAccessor:
         instructions: str,
         cache: BatchingMapProxy[str, ResponseFormat],
         response_format: Type[ResponseFormat] = str,
-        temperature: float = 0.0,
+        temperature: float | None = 0.0,
         top_p: float = 1.0,
     ) -> pd.Series:
         client: BatchResponses = BatchResponses(
@@ -205,7 +218,7 @@ class OpenAIVecSeriesAccessor:
         instructions: str,
         response_format: Type[ResponseFormat] = str,
         batch_size: int = 128,
-        temperature: float = 0.0,
+        temperature: float | None = 0.0,
         top_p: float = 1.0,
     ) -> pd.Series:
         """Call an LLM once for every Series element.
@@ -438,7 +451,7 @@ class OpenAIVecDataFrameAccessor:
         instructions: str,
         cache: BatchingMapProxy[str, ResponseFormat],
         response_format: Type[ResponseFormat] = str,
-        temperature: float = 0.0,
+        temperature: float | None = 0.0,
         top_p: float = 1.0,
     ) -> pd.Series:
         """Generate a response for each row after serialising it to JSON using a provided cache.
@@ -496,7 +509,7 @@ class OpenAIVecDataFrameAccessor:
         instructions: str,
         response_format: Type[ResponseFormat] = str,
         batch_size: int = 128,
-        temperature: float = 0.0,
+        temperature: float | None = 0.0,
         top_p: float = 1.0,
     ) -> pd.Series:
         """Generate a response for each row after serialising it to JSON.
@@ -681,7 +694,7 @@ class AsyncOpenAIVecSeriesAccessor:
         instructions: str,
         cache: AsyncBatchingMapProxy[str, ResponseFormat],
         response_format: Type[ResponseFormat] = str,
-        temperature: float = 0.0,
+        temperature: float | None = 0.0,
         top_p: float = 1.0,
     ) -> pd.Series:
         """Call an LLM once for every Series element using a provided cache (asynchronously).
@@ -848,7 +861,7 @@ class AsyncOpenAIVecSeriesAccessor:
         instructions: str,
         response_format: Type[ResponseFormat] = str,
         batch_size: int = 128,
-        temperature: float = 0.0,
+        temperature: float | None = 0.0,
         top_p: float = 1.0,
         max_concurrency: int = 8,
     ) -> pd.Series:
@@ -975,7 +988,7 @@ class AsyncOpenAIVecDataFrameAccessor:
         instructions: str,
         cache: AsyncBatchingMapProxy[str, ResponseFormat],
         response_format: Type[ResponseFormat] = str,
-        temperature: float = 0.0,
+        temperature: float | None = 0.0,
         top_p: float = 1.0,
     ) -> pd.Series:
         """Generate a response for each row after serialising it to JSON using a provided cache (asynchronously).
@@ -1040,7 +1053,7 @@ class AsyncOpenAIVecDataFrameAccessor:
         instructions: str,
         response_format: Type[ResponseFormat] = str,
         batch_size: int = 128,
-        temperature: float = 0.0,
+        temperature: float | None = 0.0,
         top_p: float = 1.0,
         max_concurrency: int = 8,
     ) -> pd.Series:

openaivec-0.13.2/src/openaivec/provider.py

@@ -0,0 +1,150 @@
+import os
+import warnings
+
+import tiktoken
+from openai import AsyncAzureOpenAI, AsyncOpenAI, AzureOpenAI, OpenAI
+
+from . import di
+from .model import (
+    AzureOpenAIAPIKey,
+    AzureOpenAIAPIVersion,
+    AzureOpenAIBaseURL,
+    EmbeddingsModelName,
+    OpenAIAPIKey,
+    ResponsesModelName,
+)
+from .util import TextChunker
+
+CONTAINER = di.Container()
+
+
+def _check_azure_v1_api_url(base_url: str) -> None:
+    """Check if Azure OpenAI base URL uses the recommended v1 API format.
+
+    Issues a warning if the URL doesn't end with '/openai/v1/' to encourage
+    migration to the v1 API format as recommended by Microsoft.
+
+    Reference: https://learn.microsoft.com/en-us/azure/ai-foundry/openai/api-version-lifecycle
+
+    Args:
+        base_url (str): The Azure OpenAI base URL to check.
+    """
+    if base_url and not base_url.rstrip("/").endswith("/openai/v1"):
+        warnings.warn(
+            "⚠️  Azure OpenAI v1 API is recommended. Your base URL should end with '/openai/v1/'. "
+            f"Current URL: '{base_url}'. "
+            "Consider updating to: 'https://YOUR-RESOURCE-NAME.services.ai.azure.com/openai/v1/' "
+            "for better performance and future compatibility. "
+            "See: https://learn.microsoft.com/en-us/azure/ai-foundry/openai/api-version-lifecycle",
+            UserWarning,
+            stacklevel=3,
+        )
+
+
+def provide_openai_client() -> OpenAI:
+    """Provide OpenAI client based on environment variables.
+
+    Automatically detects and prioritizes OpenAI over Azure OpenAI configuration.
+    Checks the following environment variables in order:
+    1. OPENAI_API_KEY - if set, creates standard OpenAI client
+    2. Azure OpenAI variables (AZURE_OPENAI_API_KEY, AZURE_OPENAI_BASE_URL,
+       AZURE_OPENAI_API_VERSION) - if all set, creates Azure OpenAI client
+
+    Returns:
+        OpenAI: Configured OpenAI or AzureOpenAI client instance.
+
+    Raises:
+        ValueError: If no valid environment variables are found for either service.
+    """
+    openai_api_key = CONTAINER.resolve(OpenAIAPIKey)
+    if openai_api_key.value:
+        return OpenAI()
+
+    azure_api_key = CONTAINER.resolve(AzureOpenAIAPIKey)
+    azure_base_url = CONTAINER.resolve(AzureOpenAIBaseURL)
+    azure_api_version = CONTAINER.resolve(AzureOpenAIAPIVersion)
+
+    if all(param.value for param in [azure_api_key, azure_base_url, azure_api_version]):
+        _check_azure_v1_api_url(azure_base_url.value)
+        return AzureOpenAI(
+            api_key=azure_api_key.value,
+            base_url=azure_base_url.value,
+            api_version=azure_api_version.value,
+        )
+
+    raise ValueError(
+        "No valid OpenAI or Azure OpenAI environment variables found. "
+        "Please set either OPENAI_API_KEY or AZURE_OPENAI_API_KEY, "
+        "AZURE_OPENAI_BASE_URL, and AZURE_OPENAI_API_VERSION."
+    )
+
+
+def provide_async_openai_client() -> AsyncOpenAI:
+    """Provide asynchronous OpenAI client based on environment variables.
+
+    Automatically detects and prioritizes OpenAI over Azure OpenAI configuration.
+    Checks the following environment variables in order:
+    1. OPENAI_API_KEY - if set, creates standard AsyncOpenAI client
+    2. Azure OpenAI variables (AZURE_OPENAI_API_KEY, AZURE_OPENAI_BASE_URL,
+       AZURE_OPENAI_API_VERSION) - if all set, creates AsyncAzureOpenAI client
+
+    Returns:
+        AsyncOpenAI: Configured AsyncOpenAI or AsyncAzureOpenAI client instance.
+
+    Raises:
+        ValueError: If no valid environment variables are found for either service.
+    """
+    openai_api_key = CONTAINER.resolve(OpenAIAPIKey)
+    if openai_api_key.value:
+        return AsyncOpenAI()
+
+    azure_api_key = CONTAINER.resolve(AzureOpenAIAPIKey)
+    azure_base_url = CONTAINER.resolve(AzureOpenAIBaseURL)
+    azure_api_version = CONTAINER.resolve(AzureOpenAIAPIVersion)
+
+    if all(param.value for param in [azure_api_key, azure_base_url, azure_api_version]):
+        _check_azure_v1_api_url(azure_base_url.value)
+        return AsyncAzureOpenAI(
+            api_key=azure_api_key.value,
+            base_url=azure_base_url.value,
+            api_version=azure_api_version.value,
+        )
+
+    raise ValueError(
+        "No valid OpenAI or Azure OpenAI environment variables found. "
+        "Please set either OPENAI_API_KEY or AZURE_OPENAI_API_KEY, "
+        "AZURE_OPENAI_BASE_URL, and AZURE_OPENAI_API_VERSION."
+    )
+
+
+CONTAINER.register(ResponsesModelName, lambda: ResponsesModelName("gpt-4.1-mini"))
+CONTAINER.register(EmbeddingsModelName, lambda: EmbeddingsModelName("text-embedding-3-small"))
+CONTAINER.register(OpenAIAPIKey, lambda: OpenAIAPIKey(os.getenv("OPENAI_API_KEY")))
+CONTAINER.register(AzureOpenAIAPIKey, lambda: AzureOpenAIAPIKey(os.getenv("AZURE_OPENAI_API_KEY")))
+CONTAINER.register(AzureOpenAIBaseURL, lambda: AzureOpenAIBaseURL(os.getenv("AZURE_OPENAI_BASE_URL")))
+CONTAINER.register(
+    cls=AzureOpenAIAPIVersion,
+    provider=lambda: AzureOpenAIAPIVersion(os.getenv("AZURE_OPENAI_API_VERSION", "preview")),
+)
+CONTAINER.register(OpenAI, provide_openai_client)
+CONTAINER.register(AsyncOpenAI, provide_async_openai_client)
+CONTAINER.register(tiktoken.Encoding, lambda: tiktoken.get_encoding("o200k_base"))
+CONTAINER.register(TextChunker, lambda: TextChunker(CONTAINER.resolve(tiktoken.Encoding)))
+
+
+def reset_environment_registrations():
+    """Reset environment variable related registrations in the container.
+
+    This function re-registers environment variable dependent services to pick up
+    current environment variable values. Useful for testing when environment
+    variables are changed after initial container setup.
+    """
+    CONTAINER.register(OpenAIAPIKey, lambda: OpenAIAPIKey(os.getenv("OPENAI_API_KEY")))
+    CONTAINER.register(AzureOpenAIAPIKey, lambda: AzureOpenAIAPIKey(os.getenv("AZURE_OPENAI_API_KEY")))
+    CONTAINER.register(AzureOpenAIBaseURL, lambda: AzureOpenAIBaseURL(os.getenv("AZURE_OPENAI_BASE_URL")))
+    CONTAINER.register(
+        cls=AzureOpenAIAPIVersion,
+        provider=lambda: AzureOpenAIAPIVersion(os.getenv("AZURE_OPENAI_API_VERSION", "preview")),
+    )
+    CONTAINER.register(OpenAI, provide_openai_client)
+    CONTAINER.register(AsyncOpenAI, provide_async_openai_client)