Mochi is a state of the art video model.
You can find all the model files for the following workflow here
diffusion_models/mochi_preview_bf16.safetensors goes in: ComfyUI/models/diffusion_models/
text_encoders/t5xxl_fp16.safetensors goes in: ComfyUI/models/text_encoders/
vae/mochi_vae.safetensors goes in: ComfyUI/models/vae/
If you have memory issues you can pick the fp8 files instead of the bf16/fp16 ones.
You can download this webp animated image and load it or drag it on ComfyUI to get the workflow.
There is also an all in one fp8 checkpoint here that contains the fp8 versions of the files in the above workflow packaged in a single checkpoint.
Note that using the fp8 files will give lower quality than using the 16 bit ones but might be faster especially if you have a GPU that supports fp8 ops.
Here is the workflow to use it:
