WorldFoundry
English中文

Mochi-1

Mochi-1 from Genmo: text-to-video; Runner parity pending.

Edit this pageGitHubMarkdown

Mochi-1 is Genmo's open text-to-video model, released as a preview checkpoint and known for strong motion quality among open models at its release. The catalog records one runnable variant, mochi-1-preview-t2v; the WorldFoundry runner path is wired for the official preview checkpoint.

No official teaser or method figure is checked into this docs tree.

Project page: genmo.ai/blog. Code: genmoai/models.

Build your run

Pipeline
Mochi1PreviewT2VPipelineworldfoundry.pipelines.video_official.pipeline_official_video
TaskInteractive Video interactive-video
Environmentworldfoundry-unified-cu128
DeviceCUDA 12.8
Runnerworldfoundry.pipeline
Loadingfrom_pretrained
Invocationnative_or_standard_invocation
worldfoundry-eval run \
  mochi-1-preview-t2v \
  --pipeline.task-profile interactive-video \
  --json

What it is for

mochi-1-preview-t2v is the recorded prompt-to-video route — not image editing and not a policy. Start here when the job is a clip from text.

Upstream describes Mochi-1 as a 10B-parameter Asymmetric Diffusion Transformer (AsymmDiT) with a dedicated video VAE. This repository records the runner wiring and the standard prompt/image/video/actions input schema; further internals are not recorded here.

Catalog tasks include text-to-video. Documented modalities: text in, video out.

Recorded evaluation entry points include VBench (Standard T2V quality suite).

Runtime contract

Optional prompt, negative_prompt, frames, fps, and height; artifacts mochi-1-preview-t2v.mp4.

CLI task profiles: interactive-video. WorldFoundry binds Mochi1PreviewT2VPipeline.

Family status: Runner parity pending.

Recorded interactive-video defaults: frames=84, fps=30, height=480, width=848, seed=12345.

No public demo reproduction script is packaged. GitHub uses genmoai while Hugging Face uses genmo; stage the preview checkpoint before running.

Running in WorldFoundry

Launch worldfoundry-eval run mochi-1-preview-t2v --pipeline.task-profile interactive-video to sample the clip. The command builder above switches --pipeline.task-profile among recorded tasks and still emits worldfoundry-eval run.

It runs in the unified environment worldfoundry-unified-cu128 (Python 3.11, CUDA 12.8).

Recorded weight repositories include genmo/mochi-1-preview. Prepare the environment and check local assets with bash scripts/inference/prepare_model_infer.sh mochi-1-preview-t2v. Add --download to fetch missing public checkpoints.

Recorded licenses include apache-2.0; status: Runner parity pending.