Voice Module - Qwen Omni TTS Configuration Guide ✨

This module interfaces with the TTS voice synthesis service provided by the Qwen multimodal model!

First, we need to copy a configuration file template from template_configs/Qwen_omni_template.toml to configs/Qwen_omni.toml.

api_key = "your_api_key" # API key
base_url = "https://dashscope.aliyuncs.com/compatible-mode/v1" # API address
model_name = "qwen-omni-turbo" # Model name
voice_character = "Chelsie" # Voice character
media_format = "wav" # Audio format, please use wav format

Configuration Item Details

api_key API Key

This is where you fill in the API key for the Qwen service. You need to go to Alibaba Cloud's DashScope Console to create and obtain your API Key.

Bailing's Warning

API Key is your important credential, please keep it safe and don't leak it to others!

base_url API Address

This is the API interface address for the Qwen service. Usually no modification is needed, just keep it as default.

model_name Model Name

This specifies the model name to use. You can choose different models according to your needs. Note, please be sure to choose a multimodal model that supports voice output! For example, qwen-omni-turbo is a good choice.

voice_character Voice Character

Here you can configure the voice character you want. Qwen provides multiple different voice character options, you can choose according to your preference.

Bailing's Tip (2025/05/22 version)

Currently, the official documentation lists supported voice characters as:

• "Cherry" （not supported by open source version models）
• "Serena" （not supported by open source version models）
• "Ethan"
• "Chelsie"

media_format Audio Format

This configures the audio format to use.

Currently, the official only supports wav format, so please modify according to the official documentation~.