Qwen-Audio-3.0-TTS-Plus Built-in Voice Narration Demo

This entry uses longanlingxin, one of the two built-in system voices of qwen-audio-3.0-tts-plus (Plus audio quality tier), to perform a real synthesis of a short Chinese narration text, demonstrating the perceptual baseline this model can achieve on the same piece of text. In terms of composition, the text is a short narration about "vocal style"; the language hint is set to zh, the sample rate is 22050 Hz, the container is mp3, and rate / pitch / volume / seed all remain at their default values. Only instruction is used to specify a calm and restrained documentary narration tone, so that any perceptual differences can be attributed to instruction control rather than parameter changes. Use cases: when selecting built-in voices for narration, podcasts, and spoken-word content, first use a short text sample to audition whether the built-in voice matches the tone of the program; it can also serve as a parameter baseline example before integrating the DashScope Speech Synthesizer. Note that this model has only two built-in voices, and the Plus tier emphasizes professional audio quality with a throughput of about 16 characters/second, making it better suited for offline auditioning and batch dubbing rather than high-throughput real-time scenarios.