# RoboVerseOrg/RoboVerse

RoboVerse: Towards a Unified Platform, Dataset and Benchmark for Scalable and Generalizable Robot Learning

Repository: https://github.com/RoboVerseOrg/RoboVerse
Canonical: https://ross.abutalabs.com/products/roboverse
Homepage: https://roboverse.wiki/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: imitation-learning, reinforcement-learning, robotics, simulation
Last push: 2026-08-24T17:12:21+00:00

## Health v2 (maintenance only)
Score: 68/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 45, longevity 36
- inputs: {"age_days": 516, "days_push": 9, "days_rel": 367, "gap_med": 20, "n_releases_24m": 2}
- flags: prerelease_only
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1810, forks 164 (observed 2026-08-28T04:05:39.472710+00:00)

## What it is
RoboVerse is a unified platform, dataset, and benchmark for scalable and generalizable robot learning, providing tasks, robots, scenes, assets, and IL/RL/VLA training workflows on top of the MetaSim simulation framework. MetaSim offers a write-once, run-anywhere interface across multiple physics engines including MuJoCo, Isaac Sim, SAPIEN, Genesis, and PyBullet.

## Use cases
- train robot policies with imitation learning across simulators
- benchmark robot learning policies on standardized tasks
- run reinforcement learning experiments in MuJoCo or Isaac Sim
- train vision-language-action models for robotics
- simulate robots across multiple physics engines with one codebase
- generate robot learning datasets with domain randomization
- evaluate generalization of robot manipulation policies

## When to choose
- you need simulator-agnostic robot learning code that runs across MuJoCo, Isaac Sim, SAPIEN, Genesis, or PyBullet
- you want a unified benchmark and dataset for comparing robot learning policies
- you are training IL, RL, or VLA policies for manipulation tasks
- you need pre-configured robots, tasks, and scene assets for robot learning research

## When to avoid
- you need a lightweight 2D simulator or simple physics sandbox
- you require a mature production robotics stack rather than an actively evolving research platform
- you cannot use a GPU or Linux/macOS environment
- you only need a single specific simulator's native tooling

## Facets
- artifact type: framework
- maturity: active
- function: simulation, machine-learning, reinforcement-learning, benchmarking, robotics
- domain: robotics, machine-learning, reinforcement-learning, simulation, artificial-intelligence
- platform: python
- tags: imitation-learning, robot-learning, mujoco, isaac-sim, sapien, genesis, pybullet, vla, benchmark, dataset, linux, macos, docker, gpu

## Member repositories
- RoboVerseOrg/RoboVerse (main) score 68

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:39.472710+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 2026-08-30T03:21:17.468450+00:00, confidence not recorded.
  - readme: https://github.com/RoboVerseOrg/RoboVerse (fetched 2026-08-28T04:05:39.472710+00:00, sha 16027e22aa62)
  - homepage: https://roboverse.wiki/ (fetched 2026-08-29T11:00:14.206124+00:00, sha 3798d4186429)
  - site_page: https://roboverse.wiki/metasim/get_started/installation.html (fetched 2026-08-29T11:00:14.217034+00:00, sha 71bc166c0032)
  - site_page: https://roboverse.wiki/FAQ (fetched 2026-08-29T11:00:14.215110+00:00, sha d77399eac0ba)
  - site_page: https://roboverse.wiki/metasim (fetched 2026-08-29T11:00:14.219918+00:00, sha 83205ba00335)
- Data as of 2026-08-30T08:39:29.467469+00:00.