# NVlabs/alpasim

AlpaSim is an open-source autonomous vehicle simulation platform designed for development and testing of end-to-end AV policies

Repository: https://github.com/NVlabs/alpasim
Canonical: https://ross.abutalabs.com/products/alpasim
Language: Python
License: Apache-2.0
License Family: permissive
Topics: autonomous-driving, autonomous-vehicles, closed-loop-simulation, computer-vision, end-to-end-driving, neural-reconstruction, nurec, nvidia, physical-ai, robotics, self-driving-car, simulation, simulator, av-simulation
Last push: 2026-08-24T14:44:08+00:00

## Health v2 (maintenance only)
Score: 75/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 75, longevity 23
- inputs: {"age_days": 334, "days_push": 9, "days_rel": 91, "gap_med": 40, "n_releases_24m": 2}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1196, forks 151 (observed 2026-08-28T04:03:57.272629+00:00)

## What it is
AlpaSim is an open-source, Python-based autonomous vehicle simulation platform for developing and testing end-to-end AV policies in closed loop. It simulates realistic sensor data (via neural renderers like NuRec), vehicle dynamics, and traffic scenarios in a modular, horizontally scalable microservices architecture.

## Use cases
- test end-to-end autonomous driving policies in closed-loop simulation
- validate new self-driving algorithms in realistic environments
- run safety analysis on edge cases and challenging driving scenarios
- benchmark and regression-test different driving models
- debug complex autonomous driving behaviors
- simulate realistic camera sensor data with neural reconstruction

## When to choose
- you need closed-loop evaluation of end-to-end AV driving policies
- you want high-fidelity, neural-reconstruction-based camera sensor simulation
- you need a hackable Python research testbed with swappable components
- you want to benchmark policies like Alpamayo, VaVAM, or Transfuser

## When to avoid
- you need a production-grade commercial AV validation toolchain with certification support
- you require non-camera sensors like lidar or radar simulation out of the box
- you want a lightweight game-engine-style driving simulator for casual use rather than ML research

## Facets
- artifact type: library
- maturity: active
- function: simulation, machine-learning, computer-vision, robotics, benchmarking, testing
- domain: autonomous-vehicles, simulation, robotics, computer-vision, machine-learning
- platform: python
- tags: autonomous-driving, closed-loop-simulation, end-to-end-driving, neural-reconstruction, sensor-simulation, self-driving-car, grpc, microservices, nvidia, research, linux, docker, gpu

## Member repositories
- NVlabs/alpasim (main) score 75

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:57.272629+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 2026-08-30T06:22:00.884240+00:00, confidence not recorded.
  - readme: https://github.com/NVlabs/alpasim (fetched 2026-08-28T04:03:57.272629+00:00, sha 6b27310152ab)
- Data as of 2026-08-30T08:39:29.467469+00:00.