To answer your question precisely, we handle all the space-separated languages and have specific tokenizers for Chinese, Japanese, Korean, Thai, and Hebrew. We plan to add more languages in the future.