add content

Author: PatrickFarley

articles/ai-services/openai/concepts/video-generation.md

@@ -1,26 +1,56 @@
+---
+title: Sora video generation overview
+description: Learn about Sora, an AI model for generating realistic and imaginative video scenes from text instructions, including safety, limitations, and supported features.
+author: PatrickFarley
+ms.author: pafarley
+manager: nitinme
+ms.service: azure-ai-openai
+ms.topic: conceptual
+ms.date: 05/22/2025
+---
 
+# Sora video generation
 
-Sora does not respond to all information provided in the prompt, but is still capable of creating realistic clips 
+Sora is an AI model from OpenAI that can create realistic and imaginative video scenes from text instructions. The model is capable of generating a wide range of video content, including realistic scenes, animations, and special effects. A variety of video resolutions and durations are supported.
 
-Users must use Sora in English or Latin script languages for best user experience 
+## Supported features
 
-Sora does not generate scenes with acts of violence but can demonstrate realistic war-like footage 
+Sora can generate complex scenes with multiple characters, diverse motions, and detailed backgrounds. The model interprets prompts with contextual and physical world understanding, enabling accurate scene composition and character persistence across multiple shots. Sora demonstrates strong language comprehension for prompt interpretation and emotional character generation. 
 
-Commonly filtered names such as Emma Watson, Obama, Biden, etc. are not filtered.  
 
-Sora does respond to prompts that can produce videos of lingerie – for example, using Japanese allowed a previously filtered prompt (in English) to be run  
 
 
-RAI: Sora has a robust safety stack including a moderation stack with prompt rewrites, content filtering, abuse monitoring, sensitive content blocking, and safety classifiers. Provenance will be added during private preview.  
 
-Sora API will support the following resolution dimensions: 480x480, 480x854, 720x720, 720x1280, 1080x1080, 1920x1080
+Sora does not respond to all information provided in the prompt, but is still capable of creating realistic clips  
+## Best practices for prompts
 
-5,10,15,20 seconds of video duration. For 1080p, max duration is 10s
+Users should write text prompts in English or Latin script languages for the best video generation performance.  
 
-context length same as gpt-image-1
+
+
+
+
+## Responsible AI
+
+Sora has a robust safety stack including a moderation stack with prompt rewrites, content filtering, abuse monitoring, sensitive content blocking, and safety classifiers.
+
+Sora does not generate scenes with acts of violence but can demonstrate realistic war-like footage  
+
+Commonly filtered names such as Emma Watson, Obama, Biden, etc. are not filtered.
 
 ## Limitations
 
+### Content quality limitations
+
+Sora may have difficulty with complex physics, causal relationships (for example, bite marks on a cookie), spatial reasoning (for example, knowing left vs. right), and precise time-base event sequencing such as camera movement.
+
+### Technical limitations
+
+Sora supports the following output resolution dimensions: 480x480, 480x854, 720x720, 720x1280, 1080x1080, 1920x1080. 
+
+Sora supports the following video durations: 5, 10, 15, and 20 seconds. The maximum duration for 1080x videos is 10 seconds.
+
+
+Max variants on 1080p =1; 720 =2; otherwise 4  
 
-Max variants on 1080p =1; 720 =2; otherwise 4
-Limiting 2 pending tasks per customer
+Limiting 2 pending tasks per customer