comfyui-workflows/models/minimax/minimax-h3-javawock2608-tts

MiniMax H3 TTS Javawock Pack

Summary

Same-day MiniMax H3 text-to-speech workflow from the javawock7618 pack. This is the most directly useful audio-focused MiniMax H3 item from today.

Model Family

• minimax

Status

• imported-only

Source

• Workflow pack page: https://huggingface.co/javawock7618/comfy-MiniMax-H3-workflows
• Workflow file: https://huggingface.co/javawock7618/comfy-MiniMax-H3-workflows/raw/main/MiniMax_int8_TTS-javano2608.1.json
• Original publisher: javawock7618
• Date imported: 2026-08-22

Developer notes

• The source README describes this as a dedicated TTS workflow.
• The graph outputs audio only, so it avoids the full video generation cost.
• It still uses the MiniMax H3 acceleration stack and a reference image input slot in the same pack structure.

Our notes

• This is the clearest same-day MiniMax H3 audio lane, but it is a helper path rather than a full audio-sync breakthrough.
• More realistic than full video on an RTX 3070, though it still looks tight.
• Good to keep if you want a local speech prototype without dragging in the whole video stack.

Required custom nodes

• ComfyUI-KJNodes
• WhatDreamsCost-ComfyUI
• ComfyUI-SolAttn_triton
• ComfyUI-Spectrum-MiniMax-H3

Required models

• minimax_h3_fl2va_pruned_int8_convrot.safetensors
• minimax_h3_video_vae_int8_convrot.safetensors
• minimax_h3_audio_vae_fp32.safetensors
• qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors
• H3/minimax_h3_fl2v_lightx2v_turbo_4step_v0.1_comfy.safetensors

Notes

• Best same-day MiniMax workflow for audio-only output.