# boundless-large-model/boundless-world-model

High-fidelity world models for general embodied intelligence, such as data engines and world  simulators.

Repository: https://github.com/boundless-large-model/boundless-world-model
Canonical: https://ross.abutalabs.com/products/boundless-world-model
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-08-19T07:22:24+00:00

## Health v2 (maintenance only)
Score: 59/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 98, release rhythm 35, longevity 11
- inputs: {"age_days": 162, "days_push": 14, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1824, forks 71 (observed 2026-08-28T04:05:41.243319+00:00)

## What it is
Boundless-World-Model (BWM) is a physically consistent, action-conditioned video world model built on Wan2.2-TI2V-5B that acts as a low-cost, high-fidelity simulator for robotic manipulation. It includes inference code, model definitions, and weights for generating action-conditioned robot manipulation videos.

## Use cases
- simulate robotic manipulation environments with a video world model
- generate action-conditioned robot manipulation videos
- build data engines for embodied AI training
- evaluate world models on the WorldArena leaderboard
- train robot learning policies with synthetic simulation data

## When to choose
- you need a high-fidelity, low-cost alternative to physical robot simulators
- you are researching action-conditioned video prediction for embodied intelligence
- you want a proven open-source world model with competition-winning results

## When to avoid
- you need training code, which is not yet released
- you lack GPU resources for large video model inference
- your use case is unrelated to robotics or embodied simulation

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, simulation, video-processing, llm-inference
- domain: artificial-intelligence, robotics, machine-learning, simulation
- platform: python
- tags: world-model, video-generation, embodied-ai, robot-manipulation, action-conditioned, simulator, wan2.2, gpu, linux

## Member repositories
- boundless-large-model/boundless-world-model (main) score 59

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:41.243319+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 2026-08-30T03:19:43.122645+00:00, confidence not recorded.
  - readme: https://github.com/boundless-large-model/boundless-world-model (fetched 2026-08-28T04:05:41.243319+00:00, sha 577a2d6e1744)
- Data as of 2026-08-30T08:39:29.467469+00:00.