It’s the former. It’s looking at audio data

So you can ask it sentiment, determine if someone is giggling, crying, laughing, can maybe even detect a condescending tone or flirtatious tone etc.

There’s an audio multimodal too

https://github.com/QwenLM/Qwen-Audio

Hugging face makes me nervous. There’s no way they’re making money now and at some point they’re going to lock down all their goodies behind a paywall