Loading a single safetensor checkpoint file for wan that was saved/combined with comfyui (non-diffuser)

As the title says, is it possible?

pipe = WanVideoPipeline.from_pretrained(
    torch_dtype=torch.bfloat16,
    device="cuda",
    model_configs=[
        ModelConfig(path=[f"{models}/aniWan2114BFp8E4m3fn_i2v480pNew.safetensors"], offload_dtype=torch.float8_e4m3fn)
    ],
    skip_download=True
)
pipe.enable_vram_management()

The above results in:

Loading models from: ['/path/to/aniWan2114BFp8E4m3fn_i2v480pNew.safetensors']

We cannot detect the model type. No models are loaded.
No wan_video_text_encoder models available.
No wan_video_dit models available.
No wan_video_vae models available.
No wan_video_image_encoder models available.
No wan_video_motion_controller models available.
No wan_video_vace models available.

Start with WanVideoPipeline.from_pretrained and ModelConfig, then trace the model-type detection that reports no models are loaded for the single checkpoint. Compare the expected Wan component layout with the combined ComfyUI safetensor input. Done means the issue's checkpoint is detected and its required Wan components load successfully, or the unsupported format is clearly documented.