✨ AI All-in-One Parsing and Scraping
Extract media metadata from almost any webpage
TIP
This feature requires MDC version 8.2.0 or later.
When conventional scraping does not work for a media item, provide the URL of a specific media information page, preferably one with complete information. AI extracts the title, description, cast, cover, and other metadata for you to review and edit.
Automatic Control and Manual Takeover
AI can recognize and click confirmation buttons on a webpage. If a CAPTCHA, login, or another action requires your input, MDC prompts you to take control of the browser. Parsing resumes automatically when you finish.
Custom Prompts
You can set site-specific prompts. Click the arrow in the lower-right corner, select Single File Scrape, then click the button to the right of the URL field. For example, you can replace characters in cover URLs or add parsing rules for a specific website.
The prompt is saved automatically for that website's domain and used again the next time you parse a page from the same site.
Before You Start
Fill in the OpenAI settings in LOCAL CONFIG or CLOUD CONFIG, and install Chrome.
For Base URL, API Key, Model, and tested providers, see the OpenAI configuration guide.
Choose a Parsing Mode
The mode switch is off by default, which uses fast mode. Turn it on to use deep mode.
| Mode | Best for | How it works |
|---|---|---|
| AI Fast Parsing | Titles, descriptions, and cast details on one page | Parses the current page with less time and token usage |
| AI Deep Parsing | Information spread across the page and related links | Opens relevant links to gather more details, using more time and tokens |
Example: Extract Media Information from a Webpage
- Launch MDC, click the arrow in the lower-right corner, select
Single File Scrape, and add the media file. - Enter the full URL of a specific media information page in the
URLfield, rather than the website's homepage. - If you need details such as cast information from related pages, turn on the mode switch for
AI Deep Parsingand start scraping. - If human input is needed, take control of the browser when MDC prompts you, then continue scraping.
- Review the returned title, description, cast, and cover. Correct any missing or inaccurate details before saving.
After starting AI parsing, expect to wait several dozen seconds. It usually finishes within a minute, depending on the response speed of the model and provider.
If the page already contains everything you need, leave the mode switch off to use AI Fast Parsing.
To process multiple movies, click CUSTOM SCRAPING at the top of the MDC window to add them in a batch and enter the corresponding information-page URL for each movie.
For a walkthrough with screenshots, see Scrape Movies That Conventional Matching Misses with AI.