generated: '2026-08-30'
method: searched
source: >-
  https://raw.githubusercontent.com/facebookresearch/habitat-sim/main/README.md,
  https://raw.githubusercontent.com/facebookresearch/habitat-lab/main/README.md
provider: AI Habitat
providerId: ai-habitat
description: >-
  AI Habitat ships no service CLI (there is no hosted API to call), but the platform's documented entry
  points ARE command-line: a compiled interactive viewer placed on PATH by the install, a Python data
  download utility invoked as a module, and example/benchmark runners. Every command below is quoted
  from the provider's own README; nothing here is inferred.
name: Habitat command-line tools
distribution: bundled with the habitat-sim / habitat-lab packages (no standalone CLI package)
install:
- method: conda
  command: conda install habitat-sim withbullet -c conda-forge -c aihabitat
  note: The README's recommended path; build flags `headless` and `withbullet` are chainable.
- method: pip-from-source
  command: pip install . --no-build-isolation
  note: Compiles from source; places the `viewer` binary on PATH.
- method: pip
  command: pip install -e habitat-lab
  note: habitat-lab core; `pip install -e habitat-baselines` adds the RL/IL baselines.
- method: docker
  command: see https://github.com/facebookresearch/habitat-lab#docker-setup
  note: Refreshed roughly once a year for the Habitat Challenge.
commands:
- group: data
  command: python -m habitat_sim.utils.datasets_download
  description: Download 3D scenes, objects and task datasets by uid into a data path.
  flags:
  - name: --uids
    description: Dataset identifier(s), e.g. habitat_test_scenes, habitat_example_objects,
      habitat_test_pointnav_dataset, replica_cad_dataset.
  - name: --data-path
    description: Destination directory for downloaded assets.
  examples:
  - python -m habitat_sim.utils.datasets_download --uids habitat_test_scenes --data-path /path/to/data/
  - python -m habitat_sim.utils.datasets_download --uids habitat_example_objects --data-path /path/to/data/
  - python -m habitat_sim.utils.datasets_download --uids replica_cad_dataset
- group: viewer
  command: viewer
  description: >-
    Compiled C++ interactive scene viewer, placed on PATH by `pip install .`. Editable/development
    builds create a ./build/viewer symlink.
  examples:
  - viewer /path/to/data/scene_datasets/habitat-test-scenes/skokloster-castle.glb
  - ./build/viewer /path/to/data/scene_datasets/habitat-test-scenes/skokloster-castle.glb
- group: viewer
  command: python examples/viewer.py
  description: Python interactive viewer, equivalent surface to the C++ viewer.
  flags:
  - name: --scene
    description: Path to a scene asset (.glb) or a named scene inside a scene dataset config.
  - name: --dataset
    description: Path to a scene_dataset_config.json.
  examples:
  - python examples/viewer.py --scene /path/to/data/scene_datasets/habitat-test-scenes/skokloster-castle.glb
  - python examples/viewer.py --dataset data/replica_cad/replicaCAD.scene_dataset_config.json --scene apt_1
- group: examples
  command: python examples/example.py
  description: Non-interactive example runner used as the smoke test for a habitat-sim install.
  flags:
  - name: --scene
    description: Path to a scene asset.
  examples:
  - python /path/to/habitat-sim/examples/example.py --scene /path/to/data/scene_datasets/habitat-test-scenes/skokloster-castle.glb
- group: training
  command: habitat-baselines (habitat_baselines run entry points)
  description: >-
    Reinforcement-learning (PPO) and imitation-learning training/eval runners shipped in
    habitat-baselines and configured with Hydra YAML config files.
  reference: https://github.com/facebookresearch/habitat-lab/blob/main/habitat-baselines/habitat_baselines/README.md
notes:
- Configuration for habitat-lab applications moved from command-line arguments to Hydra config YAML
  files in the v0.3.1 HITL work; see changelog/ai-habitat-changelog.yml.