AI Habitat

aid: ai-habitat
url: >-
  https://raw.githubusercontent.com/api-evangelist/ai-habitat/refs/heads/main/apis.yml
apis:
- aid: ai-habitat:ai-habitat
  name: AI Habitat
  tags:
  - Embodied AI
  - Simulation
  - Robotics
  - Python
  - Computer Vision
  - Reinforcement Learning
  humanURL: 'https://aihabitat.org/'
  properties:
  - url: https://aihabitat.org/
    type: Documentation
  - url: https://aihabitat.org/docs/
    type: Documentation
    title: API Documentation
  - url: https://github.com/facebookresearch/habitat-sim
    type: GitHubRepository
    title: Habitat-Sim GitHub
  - url: https://github.com/facebookresearch/habitat-lab
    type: GitHubRepository
    title: Habitat-Lab GitHub
  - url: https://pypi.org/project/habitat-sim/
    type: SDK
    title: Habitat-Sim Python Package
  - url: https://pypi.org/project/habitat-lab/
    type: SDK
    title: Habitat-Lab Python Package
  description: >-
    AI Habitat simulation framework for embodied AI research, including Habitat-Sim (high-performance 3D simulator) and Habitat-Lab (modular training library). Supports navigation, manipulation, and human-robot
    collaboration tasks across photorealistic 3D indoor environments.
name: AI Habitat
tags:
- Artificial Intelligence
- Simulation
- Embodied AI
- Robotics
- Computer Vision
- Reinforcement Learning
- Machine Learning
- Open Source
- Research
type: Contract
image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
access: 3rd-Party
created: '2025-02-17'
modified: '2026-04-19'
position: Consuming
description: >-
  AI Habitat is an open-source simulation platform from Meta AI Research for embodied AI research. It provides high-performance 3D simulated environments for training and evaluating AI agents on navigation,
  manipulation, and human-robot collaboration tasks. Habitat-Sim delivers 10,000+ FPS simulation and Habitat-Lab provides a modular library for defining tasks, training agents, and running benchmarks.
maintainers:
- FN: Kin Lane
  email: [email protected]
specificationVersion: '0.19'
common:
- name: AI Habitat GitHub Organization
  url: https://github.com/facebookresearch
  type: GitHubOrganization
  description: Meta AI Research GitHub organization hosting Habitat repositories.
- name: Habitat-Sim Repository
  url: https://github.com/facebookresearch/habitat-sim
  type: GitHubRepository
  description: High-performance 3D simulator for embodied AI research (C++/Python).
- name: Habitat-Lab Repository
  url: https://github.com/facebookresearch/habitat-lab
  type: GitHubRepository
  description: Modular library for embodied AI task definition, training, and benchmarking.
- name: Habitat Documentation
  url: https://aihabitat.org/docs/
  type: Documentation
  description: Official documentation for AI Habitat platform.
- name: Habitat Challenge
  url: https://aihabitat.org/challenge/
  type: Portal
  description: Annual Habitat Challenge competition on EvalAI for embodied AI navigation.
- name: Habitat Discussions
  url: https://github.com/facebookresearch/habitat-lab/discussions
  type: Forum
  description: Community forum for Habitat questions and support.
- name: Habitat HuggingFace
  url: https://huggingface.co/ai-habitat
  type: Portal
  description: AI Habitat datasets and models on HuggingFace.
- name: Habitat-Sim PyPI
  url: https://pypi.org/project/habitat-sim/
  type: SDK
  title: Python Package (habitat-sim)
  description: Install Habitat-Sim via pip or conda.
- name: Habitat-Lab PyPI
  url: https://pypi.org/project/habitat-lab/
  type: SDK
  title: Python Package (habitat-lab)
  description: Install Habitat-Lab via pip.
- name: PARTNR Planner
  url: https://github.com/facebookresearch/partnr-planner
  type: Tools
  description: PARTNR benchmark for human-robot collaboration using Large Planning Models with Habitat.
- type: Features
  data:
  - name: High-Performance Simulation
    description: Habitat-Sim achieves 10,000+ FPS on a single GPU and 8,000+ steps/second for robot simulation, enabling fast RL training.
  - name: Photorealistic 3D Environments
    description: Supports HM3D, MatterPort3D, Gibson, Replica, and HSSD datasets with high visual fidelity.
  - name: Physics-Enabled Simulation
    description: Bullet physics engine integration for realistic object interactions and manipulation tasks.
  - name: Robot Support via URDF
    description: Configurable robot models including Fetch mobile manipulator, Franka arm, and AlienGo quadruped.
  - name: Configurable Sensors
    description: RGB, depth, semantic, and egomotion sensors for varied agent perception configurations.
  - name: Modular Task Framework
    description: Habitat-Lab provides modular task definition, agent configuration, and benchmarking tools.
  - name: Imitation and Reinforcement Learning
    description: Built-in support for IL and RL training pipelines for embodied AI agents.
  - name: Human-Robot Collaboration
    description: Habitat 3.0 co-habitat supports humans, avatars, and robots sharing simulated environments.
  - name: Parallelizable Across Clusters
    description: Designed for large-scale distributed training across GPU clusters.
  - name: Annual Benchmark Challenge
    description: Habitat Challenge on EvalAI provides standardized evaluation of navigation and manipulation agents.
- type: UseCases
  data:
  - name: Embodied Navigation Research
    description: Train and evaluate AI agents on point-goal, object-goal, and image-goal navigation tasks in 3D environments.
  - name: Robot Manipulation Research
    description: Develop manipulation skills for pick-and-place, rearrangement, and tool use with simulated robot arms.
  - name: Human-Robot Collaboration
    description: Research human-robot teaming for household tasks using the PARTNR benchmark and Habitat 3.0.
  - name: Reinforcement Learning Training
    description: Fast simulation enables RL agents to explore millions of environment steps for policy learning.
  - name: Dataset Creation and Annotation
    description: Generate synthetic data, annotations, and demonstrations for embodied AI training datasets.
- type: Integrations
  data:
  - name: PyTorch
    description: Deep learning framework integration for neural network training and inference.
  - name: HuggingFace
    description: Datasets and models available on HuggingFace Hub at ai-habitat organization.
  - name: EvalAI
    description: Habitat Challenge evaluation hosted on EvalAI platform for standardized benchmarking.
  - name: Conda / conda-forge
    description: Conda package distribution via conda-forge and aihabitat channels.
  - name: Bullet Physics
    description: Bullet physics engine for realistic rigid-body simulation and manipulation.
  - name: ROS
    description: Robot Operating System integration for sim-to-real transfer research.