Skip to content
Get packs

CARLA

Open-source driving simulator for testing and training self-driving software.

GitHub Download 0.9.16
Stars
14.6k
Forks
4.7k
Licence
MIT
Language
C++
Last update
4 days ago
Type
App

What it is

CARLA is an open-source simulator for autonomous driving research, built to develop, train, and validate self-driving systems. It provides open digital assets such as urban layouts, buildings, and vehicles that can be used freely, and supports flexible sensor setups and environmental conditions. Researchers use it to test driving software without needing a real car on real roads.

  • Run driving scenarios with simulated vehicles, sensors, and traffic
  • Train or evaluate a driving model against simulated urban environments
  • Connect the simulator to ROS via the ROS-bridge ecosystem project
  • Run traffic scenarios with the Scenario_Runner companion project

Great for

  • Autonomous driving researchers testing perception or control software
  • Teams that want to validate driving stacks in simulated urban scenarios
  • Students and labs training imitation or reinforcement learning models on driving tasks
  • People who need configurable sensors and weather conditions for experiments

Not ideal for

  • Casual users who just want to play or try a simple game-like demo, since it requires heavy hardware and a build process
  • Anyone without a recent NVIDIA GPU and 32 GB RAM, as the README lists these as recommended specs
  • Users on Ubuntu 20.04 or Windows 10 or older, since the UE5.8 version will not work there
  • People who cannot obtain access to the Unreal Engine 5.8 fork, because building requires it
Step-by-step guide: how to use CARLAInstall, first run, examples, tips and FAQs

How to get it

  1. 1

    Clone and build from source

    You'll need: Ubuntu 22.04 or 24.04, or Windows 11 · Linked GitHub account with access to the CARLA fork of Unreal Engine 5.8 · Git · NVIDIA GPU with 16 GB or more VRAM (recommended) · Python 3 (installed by setup script on Linux if not present)

    Clone the UE5.8 development branch

    git clone -b ue58-dev https://github.com/carla-simulator/carla.git CarlaUE5

    Run the Linux setup script interactively

    cd CarlaUE5 ./CarlaSetup.sh --interactive

    Configure the build with CMake

    cmake -G Ninja -S . -B Build --toolchain=$PWD/CMake/Toolchain.cmake -DCMAKE_BUILD_TYPE=Release

    Build CARLA

    cmake --build Build
  2. 2

    Run it

    • Clone the ue58-dev branch and link your GitHub account to Epic Games for Unreal Engine access
    • Run the setup script (CarlaSetup.sh on Linux or CarlaSetup.bat on Windows) to install prerequisites and build
    • Install the Python API with cmake --build Build --target carla-python-api-install
    • Launch the editor with cmake --build Build --target launch
    • Follow the First steps guide in the documentation to run a first simulation
  3. 3

    Get the most out of it

    • Read the documentation sections on the Python API, Blueprint library, and asset catalogue
    • Use the ROS-bridge to connect CARLA to ROS for robotics-style workflows
    • Use Scenario_Runner to execute repeatable traffic scenarios
    • Use the Autonomous Driving leaderboard to validate your driving stack against a common benchmark

Downloads

Source code (.zip)ue58-dev branch · straight from GitHubRelease 0.9.16Release notes and sourceOfficial website & docscarla.org

Strengths

  • Open-source code (MIT) and open digital assets that can be used freely
  • Built specifically for autonomous driving development, training, and validation
  • Documented ecosystem of related tools such as ROS-bridge, Scenario_Runner, and a leaderboard

Trade-offs

  • Heavy hardware requirements and a complex build that requires Unreal Engine 5.8 access via an Epic Games GitHub link
  • This branch is the UE5.8 development version; the README notes significant differences from the UE4.26 version
  • Latest release 0.9.16 has no binaries, so you must build from source

How it compares

ProjectStarsTypeLevelUpdatedBest for
carla (this)14.6kAppadvanced4 days agoAutonomous driving researchers testing perception or control software
AI-For-Beginners70.5kLearning resourcebeginner13 days agoBeginners who want a structured, self-paced introduction to AI
mit-deep-learning10.5kLearning resourcebeginner2.8 years agoStudents following MIT deep learning courses
vit-pytorch25.6kLibraryintermediate5 days agoDevelopers who want to build an image classifier in PyTorch
ai-engineering-from-scratch76.3kLearning resourceintermediate4 days agoStudents who want to understand how LLMs, retrieval, and agents work internally
supervision51.5kLibraryintermediate4 days agoPython developers building apps on top of object detection or segmentation models

Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.