# NVlabs/alpamayo

NVIDIA Alpamayo 1 Nano is an open 10B reasoning VLA model for autonomous vehicles that pairs driving trajectories with Chain-of-Causation reasoning.

Repository: https://github.com/NVlabs/alpamayo
Canonical: https://ross.abutalabs.com/products/alpamayo
Language: Python
License: Apache-2.0
License Family: permissive
Topics: alpamayo, autonomous-driving, autonomous-vehicles, computer-vision, end-to-end-driving, nvidia, physical-ai, reasoning, robotics, self-driving-car, trajectory-prediction, vision-language-action, vla, world-models, chain-of-causation
Last push: 2026-08-05T19:45:32+00:00

## Health v2 (maintenance only)
Score: 59/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 96, release rhythm 35, longevity 20
- inputs: {"age_days": 287, "days_push": 28, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2005, forks 337 (observed 2026-08-28T04:06:04.666823+00:00)

## What it is
NVIDIA Alpamayo 1 is an open 10B-parameter reasoning vision-language-action (VLA) model for autonomous vehicles that pairs driving trajectory prediction with Chain-of-Causation reasoning. This repository provides setup, inference, and fine-tuning (SFT and RL) code, with future development moved to the Alpamayo Recipes hub.

## Use cases
- run a reasoning VLA model for autonomous driving trajectory prediction
- generate chain-of-causation explanations for driving decisions
- fine-tune a driving model with supervised fine-tuning
- post-train a driving policy with reinforcement learning
- research end-to-end self-driving with vision-language-action models
- evaluate world models for physical AI driving tasks

## When to choose
- you need an open reasoning VLA model for autonomous driving research
- you want trajectory prediction paired with causal reasoning traces
- you need a base model for SFT or RL post-training on driving data

## When to avoid
- you need the latest features and active support - use the newer Alpamayo Recipes repository
- you need a production-ready self-driving stack rather than a research model
- you lack the GPU hardware required for a 10B-parameter model

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, computer-vision, llm-inference, llm-training, simulation
- domain: autonomous-vehicles, robotics, deep-learning, artificial-intelligence, computer-vision
- platform: python
- tags: vision-language-action, autonomous-driving, trajectory-prediction, reasoning, world-models, chain-of-causation, physical-ai, self-driving, vla, end-to-end-driving, gpu, linux

## Member repositories
- NVlabs/alpamayo (main) score 59

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:04.666823+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 2026-08-30T03:01:29.766921+00:00, confidence not recorded.
  - readme: https://github.com/NVlabs/alpamayo (fetched 2026-08-28T04:06:04.666823+00:00, sha 59dea0494d95)
- Data as of 2026-08-30T08:39:29.467469+00:00.