acrolinx

Author: PatrickFarley

articles/ai-services/openai/concepts/content-filter-streaming.md

@@ -1,6 +1,6 @@
 ---
 title: Content Filter Streaming in Azure OpenAI
-description: Learn about content filter streaming options in Azure OpenAI, including default and asynchronous filtering modes, and their impact on latency and Guidelines & controls performance.
+description: Learn about content filter streaming options in Azure OpenAI, including default and asynchronous filtering modes, and their impact on latency and performance.
 author: PatrickFarley
 manager: nitinme
 ms.service: azure-ai-openai
@@ -16,7 +16,7 @@ This guide describes the Azure OpenAI content streaming experience and options.
 
 ## Default filtering behavior
 
-The content filtering system is integrated and enabled by default for all customers. In the default streaming scenario, completion content is buffered, the content filtering system runs on the buffered content, and – depending on the content filtering configuration – content is either returned to the user if it doesn't violate the content filtering policy (Microsoft's default or a custom user configuration), or it’s immediately blocked and a content filtering error is returned instead. This process is repeated until the end of the stream. Content is fully vetted according to the content filtering policy before it's returned to the user. Content isn't returned token-by-token in this case, but in “content chunks” of the respective buffer size.
+The content filtering system is integrated and enabled by default for all customers. In the default streaming scenario, completion content is buffered, the content filtering system runs on the buffered content, and – depending on the content filtering configuration – content is either returned to the user if it doesn't violate the content filtering policy (Microsoft's default or a custom user configuration), or it is immediately blocked and a content filtering error is returned instead. This process is repeated until the end of the stream. Content is fully vetted according to the content filtering policy before it's returned to the user. Content isn't returned token-by-token in this case, but in "content chunks" of the respective buffer size.
 
 ## Asynchronous filtering
 
@@ -28,7 +28,7 @@ Customers must understand that while the feature improves latency, it's a trade-
 
 **Content filtering signal**: The content filtering error signal is delayed. If there is a policy violation, it’s returned as soon as it’s available, and the stream is stopped. The content filtering signal is guaranteed within a ~1,000-character window of the policy-violating content. 
 
-**Customer Copyright Commitment**: Content that is retroactively flagged as protected material may not be eligible for Customer Copyright Commitment coverage. 
+**Customer Copyright Commitment**: Content that is retroactively flagged as protected material might not be eligible for Customer Copyright Commitment coverage. 
 
 To enable Asynchronous Filter in [Azure AI Foundry portal](https://ai.azure.com/), follow the [Content filter how-to guide](/azure/ai-services/openai/how-to/content-filters) to create a new content filtering configuration, and select **Asynchronous Filter** in the Streaming section.
 
@@ -92,7 +92,7 @@ data: {
 
 ### Annotation message
 
-The text field will always be an empty string, indicating no new tokens. Annotations will only be relevant to already-sent tokens. There may be multiple annotation messages referring to the same tokens.  
+The text field is always an empty string, indicating no new tokens. Annotations are only relevant to already-sent tokens. There might be multiple annotation messages referring to the same tokens.  
 
 `"start_offset"` and `"end_offset"` are low-granularity offsets in text (with 0 at beginning of prompt) to mark which text the annotation is relevant to. 
 
@@ -181,4 +181,4 @@ data: [DONE]
 ```
 
 > [!IMPORTANT]
-> When content filtering is triggered for a prompt and a `"status": 400` is received as part of the response there will be a charge for this request as the prompt was evaluated by the service. Due to the asynchronous nature of the content filtering system, a charge for both the prompt and completion tokens will occur. [Charges will also occur](https://azure.microsoft.com/pricing/details/cognitive-services/openai-service/) when a `"status":200` is received with `"finish_reason": "content_filter"`. In this case the prompt did not have any issues, but the completion generated by the model was detected to violate the content filtering rules which results in the completion being filtered.
+> When content filtering is triggered for a prompt and a `"status": 400` is received as part of the response there will be a charge for this request as the prompt was evaluated by the service. Due to the asynchronous nature of the content filtering system, a charge for both the prompt and completion tokens occurs. [Charges will also occur](https://azure.microsoft.com/pricing/details/cognitive-services/openai-service/) when a `"status":200` is received with `"finish_reason": "content_filter"`. In this case, the prompt didn't have any issues, but the completion generated by the model was detected to violate the content filtering rules, which results in the completion being filtered.