Mureka AI Mureka V7.6 Recognize Song (Official Edition)

Mureka's V7.6 "Recognize Song" node is the closest thing ComfyUI has to a built-in Shazam. You feed it a song, it tells you what it is - and it runs on Mureka's V7.6 recognition model, an "Official Edition" API node in BizyAirPlus. It's the node you don't think you need until you're staring at an audio file with no metadata and a deadline.

What it is

A song-identification node. Where its sibling Mureka_Official_S2T describes a song's genre and structure, this one identifies it - matching the audio against Mureka's catalog of known tracks. Different job, different model, same shape of input.

How it works

The audios input takes the song to recognize - a hosted URL, consistent with the other Mureka audio nodes. The V7.6 recognition model fingerprint-matches the clip and returns the identification as text in the result output. It's an output node; nothing needs to hang off it, though you can capture result into anything that wants a STRING.

The inputs and outputs

• audios - required. The song to recognize.
• outputs - result (the identification text) and request_id.

api_config and skip_error are the standard pack options. That's the entire surface - this node is deliberately single-purpose.

Installing

Via ComfyUI Manager (search "BizyAirPlus") or manually:

cd ComfyUI/custom_nodes
git clone https://github.com/siliconflow/BizyAirPlus.git
cd BizyAirPlus
python -m pip install -r requirements.txt

Restart, toggle the BizyAirPlus button ON, and drop in your BizyAir API key from bizyair.ai.

Where people get burned

Same URL gotcha as every Mureka audio node: local files don't work, hosted URLs do. More interestingly, set your expectations on the recognition - V7.6 is a fingerprint matcher, so a heavily remixed or low-quality recording may come back unknown. And there's an overlap trap with the S2T Describe node: if you actually want a description of style, you used the wrong node. This one answers "what song is this," not "what's it like."