Nodes/ComfyUI-Kie-API/KIE Suno Music (Basic)
ComfyUI Node

KIE Suno Music (Basic)

KIE Suno Music (Basic)

By gateway·Created 9 months ago·Updated 4 months ago· 39
KIE Suno Music (Basic)
    • audio_1
    • audio_2
    • data
    • image_1
    • image_2
    title
    style
    prompt
    custom_modetrue
    instrumentaltrue
    modelV4_5
    negative_tags
    vocal_gendermale
    logtrue

    Suno Music (Basic) is the node you use when you want a track and don't want to think about it. It's the stripped-down sibling of the Advanced variant in the same pack: same Suno backend, same two-tracks-and-two-covers output, same models - but only the handful of fields you actually touch. For "give me a mood, roughly this genre, surprise me," it's the right tool, and honestly the better first stop if you're new to the pack's audio side.

    The name is the pitch. Where Advanced exposes style/weirdness/audio weights for fine-tuning a sound, Basic just wants a title, a style, a prompt, and a model - and it does the rest.

    How it works

    Suno on KIE runs through the /api/v1/generate endpoint and /generate/record-info polling rather than the pack's usual createTask job flow. The node watches the track pass through PENDING, TEXT_SUCCESS, and FIRST_SUCCESS states - you'll literally see the lyrics get written before the audio generates. When it lands, you get two AUDIO outputs and two IMAGE covers, plus the full record JSON.

    The inputs that matter

    • title - required in custom mode (which is on by default). Keep it short; it's a track title.
    • style - the genre and vibe text. This is the highest-leverage field: "dreamy lo-fi hip hop, rain sounds, vinyl crackle" beats a vague "chill". Multi-line allowed.
    • prompt - lyrics when instrumental is off, otherwise context. Max length depends on model (3000 chars for V4, 5000 for newer ones), and the node validates that for you.
    • instrumental - defaults to true. Flip it off and the prompt becomes your vocal lyrics.
    • model - V4_5 default, with V4, V4_5PLUS, V4_5ALL, and V5 available.
    • vocal_gender and negative_tags - the only extras, and both optional. Gender only applies in custom mode.

    Outputs: audio_1, audio_2 (AUDIO), image_1, image_2 (IMAGE), and data (record JSON). All five come out of one run - that's two full tracks for the price of one generation call.

    Installing

    The whole KIE pack installs once - you don't install Basic separately from Advanced. ComfyUI Manager: search "ComfyUI Kie API". Or:

    cd ComfyUI/custom_nodes
    git clone https://github.com/gateway/ComfyUI-Kie-API
    # restart ComfyUI
    

    No model downloads. The API key goes in config/kie_key.txt inside the pack folder (copy from kie_key.example.txt).

    Where people get burned

    First: two tracks per run means two charges per run. Basic feels cheap until you batch ten generations. Second: Suno rejects lyrics containing sensitive words with a SENSITIVE_WORD_ERROR - that's a prompt problem, not a node problem, so reword. Third: generation is slow; the node polls patiently, so give it time before you assume failure. If something does die, the console log shows the state transitions and the returned data JSON holds the record. And if Basic is too coarse and you want Suno's style/audio weights exposed, that's exactly what the Advanced variant adds - same install, no setup work.

    Categorykie/api

    Inputs (9)

    NameTypeDefaultDescription
    titleSTRING
    styleSTRING
    promptSTRING
    custom_modeBOOLEANtrue
    instrumentalBOOLEANtrue
    modelCOMBOV4_55 options: V4, V4_5, V4_5PLUS, V4_5ALL, V5
    negative_tagsoptSTRING
    vocal_genderoptCOMBOmale2 options: male, female
    logoptBOOLEANtrue

    Outputs (5)

    NameTypeDescription
    audio_1AUDIO
    audio_2AUDIO
    dataSTRING
    image_1IMAGE
    image_2IMAGE