Instructions
This workflow is modified based on the official MiniMax H3 All-in-One Reference workflow, with added automatic prompt nodes and prompt templates.
The automatic prompt node can only analyze a single image. Therefore, when using multi-image reference, it is recommended to use the prompt template on web platforms such as GPT and Gemini, or with agents like Codex and Claude.
(Language model download link: https://huggingface.co/AVert888/Qwen3.5-9B-mmproj-q8_0-GGUF)
Resolution is adjusted by modifying numerical values, with detailed explanations provided within the workflow.
H3 model download link: https://huggingface.co/Comfy-Org/MiniMax-H3
Available model variants and VRAM requirements:
minimax_h3_ref2va_pruned_int8_convrot.safetensors: 12GB VRAMminimax_h3_ref2va_int8_convrot.safetensors: 16GB VRAMminimax_h3_ref2va_bf16.safetensors: 16GB+ VRAM
Please select according to your hardware configuration.
Runtime reference (based on RTX 5060 Ti 16GB + 32GB RAM):
Generating a 10-second 480p video with
minimax_h3_ref2va_int8_convrot.safetensorstakes approximately 950 seconds.It is recommended to try the free online version first. If you are satisfied with the results, you may deploy it locally.
Free online workflow: https://www.runninghub.ai/post/2079133532412309506/?inviteCode=rh-v1443
Description
FAQ
Comments (2)
An English version would be nice, so I know what each of the nodes is doing.
Hi, I've had to bypass the Llama nodes because I keep getting this error, even though I downloaded the same models as you and changed Python and reinstalled that node package. Do you know why this might be?
[llama-cpp_vlm] Loading model...
[llama-cpp_vlm] Loading clip: mmproj-Q8_0.gguf
[llama-cpp_vlm] Loading model: Qwen3.5-9B-Q8_0.gguf
[llama-cpp_vlm] n_gpu_layers = -1
[ERROR] !!! Exception during processing !!! Image input detected, but the loaded model is not configured with a mmproj module.
[ERROR] Traceback (most recent call last):
File "C:\Users\***\ComfyUI-Installs\ComfyUI\ComfyUI\execution.py", line 545, in execute
output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\***\ComfyUI-Installs\ComfyUI\ComfyUI\execution.py", line 344, in get_output_data
return_values = await asyncmap_node_over_list(prompt_id, unique_id, obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\***\Documents\ComfyUI\custom_nodes\comfyui-lora-manager\py\metadata_collector\metadata_hook.py", line 171, in async_map_node_over_list_with_metadata
results = await original_map_node_over_list(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\***\ComfyUI-Installs\ComfyUI\ComfyUI\execution.py", line 318, in asyncmap_node_over_list
await process_inputs(input_dict, i)
File "C:\Users\***\ComfyUI-Installs\ComfyUI\ComfyUI\execution.py", line 306, in process_inputs
result = f(**inputs)
^^^^^^^^^^^
File "C:\Users\***\Documents\ComfyUI\custom_nodes\ComfyUI-llama-cpp_vlm\nodes.py", line 547, in process
raise ValueError("Image input detected, but the loaded model is not configured with a mmproj module.")
ValueError: Image input detected, but the loaded model is not configured with a mmproj module.