Xiaoxiao is the most widely used Mandarin Chinese female neural voice on Microsoft Azure. Five variants are listed below — standard, dialects (13 Chinese dialects), multilingual (91 locales), and two Dragon HD Flash Latest versions with context-aware delivery. Click the preview button on each card to hear a sample.
Xiaoxiao (晓晓) is a warm, well-rounded and animated female Mandarin Chinese neural voice from Microsoft Azure. It is one of the most widely used Chinese TTS voices for short-video narration, audiobooks, customer-service prompts and voice assistants.
Five Xiaoxiao variants are listed here: the 48kHz standard Xiaoxiao voice with 20 emotional styles, Xiaoxiao Dialects covering 13 Chinese dialects, Xiaoxiao Multilingual covering 91 locales, and two Dragon HD Flash Latest versions with context-aware delivery and 15/20 styles. Click the preview button on each card to hear a sample, or open the main text-to-speech tool to convert your own text.
Common use cases for Xiaoxiao
Chinese short-video voiceovers for Douyin, Bilibili, Xiaohongshu
Audiobooks and long-form narration using Xiaoxiao2 (story-telling style)
Customer-service prompts and IVR using the customerservice style
Chinese dialect videos using the Xiaoxiao Dialects variant
Mixed Chinese-English content using Xiaoxiao Multilingual or the Dragon HD variants
The interface is in Chinese; the input box accepts Chinese, English or any other supported language text. Pick any voice whose name starts with “Xiaoxiao” in the voice dropdown.
Xiaoxiao is a female Mandarin Chinese neural voice from Microsoft Azure. All 5 Xiaoxiao variants (standard, dialects, multilingual, and two Dragon HD Flash Latest versions) are female voices with a warm, well-rounded and animated tone.
Which Xiaoxiao variant should I use for audiobooks?
+
Use Xiaoxiao2 Dragon HD Flash Latest for audiobooks and long-form narration. It supports 20 emotional styles including story-telling, poetry-reading, sentimental and tired, plus context-aware delivery. The standard Xiaoxiao voice (48kHz, 274 words/min) is also suitable for audiobooks and chat scenarios.