zhenzhen-mureka-bgm-lowprice

You need a 30-second bed under a video, you don't want to license a track, and you definitely don't want to compose one. Comfly_mureka_bgm_lowprice ("zhenzhen-mureka-bgm-lowprice") is the node that hands you exactly that: pick mureka-v8-bgm or mureka-v9-bgm, describe the vibe ("lo-fi hip hop, 85 BPM, warm bass"), and get back generated background music as AUDIO. Mureka is Moonshot/Kimi's music model family, and the "-bgm" variants are the instrumental-focused members - built for exactly this "bed track" job, with less of the lyric-driven full-song baggage.

It's a "lowprice" node in the T8mars/Comfyui-zhenzhen pack, the "Comfly" API-wrapper family that calls the author's paid reseller service. That means it uses the cheaper China-facing endpoint (api.seedance.nz) and takes its key from the shared zhenzhen-seedance2-low-price-api-settings node through the api_config input, rather than a per-node key field. Submit → poll → download, and the pack's changelog shows the v8/v9 BGM models tracking Mureka's releases.

The inputs that matter

• model - required: mureka-v8-bgm or mureka-v9-bgm.
• prompt - required. This is the whole creative input, so be specific about genre, tempo, mood, and instrumentation.
• instrumental_id - optional; if you have an instrumental track id you want to build on or reference, this is where it goes.
• n - 1 to 3 tracks per run. Generating a couple of candidates in one call is the sensible workflow when the first one isn't right.
• skip_error - fail soft for batch runs, and a cache-control seed (not sent to the API).

Outputs

• audios - an AUDIO list (note: this one's a list, one entry per track). Feed it into an audio-preview or a save node.
• audio_urls - the hosted files as a string.
• task_id - for the site's async task list.
• response - the raw payload.

Installing it

cd ComfyUI/custom_nodes
git clone https://github.com/T8mars/Comfyui-zhenzhen
# restart ComfyUI

Or ComfyUI Manager → "Comfyui-zhenzhen". Then a zhenzhen-seedance2-low-price-api-settings node with your key, wired into api_config.

The honest part

For a working video pipeline, "BGM node that just works" is quietly valuable - nobody writes a blog post about it, but every edit needs music. The BGM variants are the right tool here: they skip the vocals problem entirely, so you don't get an AI song belting over your dialog. The usual pack caveats apply - your prompt goes through a reseller, cost is per call (use n wisely), and if you're generating scores by the hour, a local model is still cheaper at volume. For one-off beds under a video that needs to ship today, this is the fastest path in the pack.