Get the parsing process and results of Parse.
Parse returns structured results for downstream use. This page explains how to read parsed content, verify it against the original file, and hand the results to Agent applications.
If you need additional position information or want to adjust what is retained, read Configure Document Parsing.
Audio and video results are organized by content and time position. For audio, focus on speech; for video, also check whether visuals and text descriptions correspond.
Prerequisites and Entry Points
| Item | Requirement |
|---|---|
| Task | A document parsing task that has completed successfully |
| Current result | After the run completes, view it in the "Document Parsing" workspace |
| Historical results | Open the result of the corresponding task from Task Records |
View Text-and-Image Results
Choose a Result View
| View | Purpose | Focus on |
|---|---|---|
| Markdown | Read parsed content continuously | Headings, paragraphs, tables, and image descriptions |
| JSON | View structured content | Content and related information that programs need to read |
Compare Against the Original File
- In the original file, select the page you want to verify.
- Find the corresponding heading, paragraph, or table in the result.
- Check key content item by item, especially data that the business will use directly.
| Content | How to verify |
|---|---|
| Body text | Check that headings, paragraph order, and content are complete |
| Tables | Check headers, row-column relationships, amounts, dates, and units |
| Image descriptions | Compare with the original images and check whether the descriptions match the relationships in the images |
| Content positions | If the application needs to locate the original text, check the position information in the result |
View Audio and Video Results
Audio and video results are organized by content and time position. For audio, focus on speech; for video, also check whether visuals and text descriptions correspond.
Choose a Video Result View
| Page view | What it tells you |
|---|---|
| Story overview | Overall video content, main characters, and other information |
| Parsed segment results | Speech, visuals, and related descriptions for each time period |
| Story segments | Segments divided by content and their topics |
| Story summary | Main points in the video |
Locate and Verify Segments
- Open "Parsed Segment Results".
- Find the segment you want to view on the timeline.
- Click the segment's time position to jump the player to the corresponding content.
- Play the original file and verify the text, visuals, and surrounding context.
Use the Results
| Usage | Action |
|---|---|
| Manual viewing or copying | Use the copy button in the console result area to take text or structured content |
| Hand to your own Agent application | Get task results via REST API Integration and read the returned content or result files |
| Continue extracting specified fields from text-and-image results | Select a reusable parsed document in the Field Extraction Console |
Related Guides
- Document Parsing Overview: Review document parsing capabilities and the complete task flow.
- Configure Document Parsing: Check file information, parsing scope, enhancement options, and output forms.