Nodes/ComfyUI-GGUF-Loader/Empty MiniMax H3 AV Latent (Batch) ⚡
ComfyUI Node

Empty MiniMax H3 AV Latent (Batch) ⚡

Joint video+audio latent for MiniMax H3 with a batch size. Duration snaps to the model's 17k+5 frame grid at 24 fps. For batch_size > 1, route the model through MiniMax H3 Batch Patch.

By ChrisColeTech·Created 13 days ago·Updated about 3 hours ago· 6
Empty MiniMax H3 AV Latent (Batch) ⚡
    • LATENT
    width1344
    height768
    length124
    batch_size1
    Category🤖 CCTech/MiniMax H3

    Inputs (4)

    NameTypeDefaultDescription
    widthINT134432–16384
    heightINT76832–16384
    lengthINT1245–3600Frame count at 24 fps, snapped up to the model's 17k+5 grid (124 = ~5s; trained range is ~124-362, longer is untested).
    batch_sizeINT11–64Videos per job. Each gets distinct noise from the sampler seed. Combine with MiniMax H3 Batch Patch: the DiT itself only accepts batch 1, so the patch node is required for values above 1.

    Outputs (1)

    NameTypeDescription
    LATENTLATENT