MiniMax-H3
This demo is a trimmed MiniMax-H3 module within the VideoGen module of the Comfy web UI. Supports High-Level MCP 🤖
MiniMax H3 Video Generation
💡 Tip: No image for T2VA; upload First Frame for I2VA; upload First Frame & Last Frame for FL2VA. Under the default ZeroGPU Duration (s) of 120s, tasks with 544p resolution, 20 steps, and 5s duration pass successfully.
💡 Tip: When downloading from Civitai, please use the Version ID, not the Model ID. You can find the Version ID in the URL (e.g., civitai.com/models/123?modelVersionId=456) or under the model's download button. When downloading from Hugging Face, please use the format: repo_id/filename.extension or repo_id/folder_path/filename.extension (e.g., Comfy-Org/MiniMax-H3/loras/minimax_h3_fl2v_turbo_4step_v1.0_768p_comfyui_bf16.safetensors or Comfy-Org/MiniMax-H3/loras/minimax_h3_ref2v_turbo_4step_v0.1_comfyui_bf16.safetensors).
MiniMax H3 Reference Video Generation
💡 Tip: Supports uploading up to 9 reference images, 3 reference videos with audio, and 3 reference audio files. Under the default ZeroGPU Duration (s) of 120s, tasks with 544p resolution, 20 steps, 5s duration, and 3 reference images pass successfully.
💡 Tip: When downloading from Civitai, please use the Version ID, not the Model ID. You can find the Version ID in the URL (e.g., civitai.com/models/123?modelVersionId=456) or under the model's download button. When downloading from Hugging Face, please use the format: repo_id/filename.extension or repo_id/folder_path/filename.extension (e.g., Comfy-Org/MiniMax-H3/loras/minimax_h3_fl2v_turbo_4step_v1.0_768p_comfyui_bf16.safetensors or Comfy-Org/MiniMax-H3/loras/minimax_h3_ref2v_turbo_4step_v0.1_comfyui_bf16.safetensors).