# OpenMotionLab/MotionGPT

[NeurIPS 2023] MotionGPT: Human Motion as a Foreign Language, a unified motion-language generation model using LLMs

Repository: https://github.com/OpenMotionLab/MotionGPT
Canonical: https://ross.abutalabs.com/products/motiongpt
Homepage: https://motion-gpt.github.io
Language: Python
License: MIT
License Family: permissive
Topics: 3d-generation, chatgpt, gpt, language-model, motion, motion-generation, text-driven, text-to-motion, motiongpt, multi-modal
Last push: 2025-07-01T18:22:15+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 29, release rhythm 8, longevity 83
- inputs: {"age_days": 1170, "days_push": 428, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1961, forks 144 (observed 2026-08-28T04:05:59.422813+00:00)

## What it is
MotionGPT is a unified motion-language model that treats 3D human motion as a foreign language by converting motion into discrete motion tokens and modeling them alongside text with an LLM. It supports multiple motion tasks including text-to-motion generation, motion captioning, motion prediction, and motion in-betweening.

## Use cases
- generate 3d human motion animations from text descriptions
- generate text captions describing human motion sequences
- predict future motion frames from past motion
- fill in missing motion between keyframes
- answer questions about motion sequences with an llm
- research multimodal motion-language models

## When to choose
- you need text-to-motion or motion captioning in a research pipeline
- you want a unified model handling multiple motion tasks
- you want a reproducible NeurIPS 2023 baseline for motion-language modeling

## When to avoid
- you need production-ready real-time animation tooling
- you lack GPU resources for large model inference
- you need game-engine-native motion synthesis without Python

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, llm-inference, nlp, data-generation
- domain: machine-learning, deep-learning, large-language-models, computer-vision, artificial-intelligence
- platform: python, cross-platform
- tags: text-to-motion, motion-generation, motion-capture, multimodal, motion-tokens, neurips-2023, research, pytorch, linux, macos

## Member repositories
- OpenMotionLab/MotionGPT (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:59.422813+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:05:37.029131+00:00, confidence not recorded.
  - readme: https://github.com/OpenMotionLab/MotionGPT (fetched 2026-08-28T04:05:59.422813+00:00, sha 962920957b7e)
  - homepage: https://motion-gpt.github.io (fetched 2026-08-29T10:45:32.384542+00:00, sha 04f2c380a31f)
- Data as of 2026-08-30T08:39:29.467469+00:00.