OCR perform badly in Chinese content

[hetailang]

Describe the bug
I used the UI of llamaParse. When receive Chinese content, llamaParse output meanless content.
Files

Job ID
d873ef1c-3fb0-42df-b5e7-1482d0ee25f3

Screenshots
The output is shown as bellow

it only recognize english characters like https://movie.douban.com/subject/35351365

Options
I set OCR language field as ch_sim and leave other fields as default.

Additional context
without setting gpt4o_mode=True, the result still makes no sense using in python, so I guess the problem comes from the ocr used in llamaparse