Web & Social
In addition to text and documents, MindLM also supports various input methods including web links, image recognition, Twitter threads, and e-books.
Web Links
Paste a public web page URL, and the system will automatically extract the main content and generate a mind map.
Character Limits
| Membership Tier | Web Page Character Limit |
|---|---|
| Free | 10,000 characters |
| Pro | 100,000 characters |
| Max | 300,000 characters |
Limitations and Notes
The following types of web pages cannot be properly parsed:
- Pages requiring login to access
- Websites with strong anti-scraping measures
- Single-page applications with heavy dynamic rendering (some content may not load completely)
- Internal network or private domain addresses (only public URLs are supported)
Web parsing automatically filters common interfering elements (such as navigation bars, ads, footers), but for pages with messy content, it's recommended to copy the core text first and use text mode for generation.
Image OCR
Upload an image, and the system will automatically recognize the text in the image and generate a mind map.
Supported Formats
PNG, JPG, JPEG, WEBP
Size Limit
The limit per file is 10MB.
Image OCR is available to all membership tiers. Recognition languages are primarily Simplified Chinese and English. The clearer the image and more standard the text, the better the recognition results.
Twitter Threads
Twitter thread feature is only available to Pro / Max members.
Paste a Twitter (X) post or thread link, and the system will extract the content and generate a mind map. Ideal for summarizing trending discussions or knowledge threads.
E-books
Upload e-book files, and the system will extract key information and generate a knowledge mind map. Perfect for quickly outlining book highlights.
E-book feature is available to all membership tiers. It's recommended to upload single books rather than entire libraries for better single-topic results.
Common Questions
QWhat if web parsing only captures partial content?
You can manually copy the core content of the web page and paste it into the text input box. Using text mode provides more stable generation results.
QWhat if image OCR text recognition is inaccurate?
Try uploading a higher resolution image, or crop the image to the text area and upload again.