According to the official Alibaba Cloud website, the BaiLian large model service platform is set to reduce pricing for certain speech series models at its domestic service nodes in Beijing and Singapore. These changes will take effect starting at 00:00:00 Beijing Time on September 22, 2026. The models affected by this price adjustment include the qwen-audio-3.1-tts-flash and qwen-audio-3.0-realtime-flash.

For the qwen-audio-3.1-tts-flash model, the input price will be revised to 1.5 yuan per million tokens, while the output price will be set at 12 yuan per million tokens. Meanwhile, for the qwen-audio-3.0-realtime-flash model, the pricing will be adjusted as follows: the input text price will be 1.5 yuan per million tokens, the input audio price will be 6 yuan per million tokens, the output text price will be 4.5 yuan per million tokens, and the combined output text + audio price will be 12 yuan per million tokens.

These new rates will be automatically applied at the designated effective time, and users will not be required to take any action for the changes to take effect.