問題文
In what two formats can Content Understanding deliver its final structured output?
選択肢
- Only a plain text transcript with no structure
- Only a binary format that must be read through the SDK, which keeps the confidence scores and the grounding references attached to each value so that they cannot be separated from the value they describe
- Markdown for search and retrieval scenarios, or structured JSON matching a defined schema for automation and analytics
- An audio summary and a video highlight reel