You can translate documents in most mainstream formats, including PDF, Word, Excel, PowerPoint, SRT subtitles, ePub e-books, RAR, ZIP, etc. Meanwhile, we are continuously iterating to support more formats.

Large language models are slower than traditional machine translation. Depending on the amount of text to be translated in your document, it mostly won't take more than a minute.

Your documents are translated by the best large language model currently available, representing the best level of AI translation.

If you are a newly registered user, you will have several free translation opportunities. Once these free attempts are used up, the translation feature will switch to a paid model. A significant portion of the cost comes from the computing power expense of large language models. When translating documents, we need to repeatedly call the content generation interface of large language models. To pursue better translation effects, we may improve our algorithms and increase the number of calls.

Yes, our software will extract text for translation based on the PDF document structure information as much as possible, and then put the translation back into its original position, preserving the document's original layout. Although it's not perfect, it's probably the best effect currently available. You can check out the examples displayed on our official website. Meanwhile, we are continuously improving our algorithms to ensure more accurate layout preservation.

For your convenience to retrieve translation history, we will store your documents privately on our servers. The storage limit for regular users is 3 days, while subscribed users can store for 7 to 30 days.