The following table lists the answer snippet model supported for data from different types of sources.
Note that currently, only plain text data extraction is supported for answer snippets, and data from tables or images are not extracted.
| Sources | File Type | Extractive Model | Generative Model |
| Web | Static HTML | Yes | Yes |
| File | Yes | Yes | |
| docx | No | Yes | |
| ppt | No | Yes | |
| xlsx | No | No | |
| txt | No | Yes | |
| png | No | No | |
| jpeg | No | No | |
| jpg | No | No | |
| FAQ | Yes | Yes | |
| Structure Data | json | Yes | Yes
Note: Only the fields, “title” & “Content”, if available, are used for generating snippets. |
| Connectors | |||
| Confluence Server | Knowledge Articles | Yes | Yes |
| Confluence Cloud | Knowledge Articles | Yes | Yes |
| Service Now | Knowledge Articles | Yes | Yes |
| Sharepoint | Pages/Articles (aspx) | Yes | Yes |
| docx | No | Yes | |
| xlsx | No | No | |
| pptx | No | Yes | |
| Yes | Yes | ||
| doc | No | Yes | |
| xls | No | No | |
| ppt | No | Yes | |
| html | Yes | Yes | |
| txt | No | Yes | |
| csv | No | No | |
| sql | No | No | |
| xhtml | Yes | Yes | |
| Zendesk | Tickets/json | No | No |
| Knowledge Articles | Yes | Yes | |
| Google drive | docx | No | Yes |
| doc | No | Yes | |
| xlsx | No | No | |
| ppt | No | Yes | |
| pptx | No | Yes | |
| No | Yes | ||
| txt | No | Yes | |
| html | No | Yes | |
| Azure Storage | txt | No | Yes |
| Yes | Yes | ||
| rtf | No | Yes | |
| msword | No | Yes | |
| docx | No | Yes | |
| doc | No | Yes | |
| xlsx | No | No | |
| xls | No | No | |
| ppt | No | Yes | |
| pptx | No | Yes | |
| Salesforce | Knowledge articles | No
Extractive Snippets are supported if there is ‘html’ content. |
Yes |
| Oracle Knowledge | Knowledge articles | Yes | Yes |
| Dropbox | docx | No | Yes |
| doc | No | Yes | |
| xlsx | No | No | |
| ppt | No | Yes | |
| pptx | No | Yes | |
| Yes | Yes | ||
| txt | No | Yes | |
| html | Yes | Yes |