# NVlabs/VILA

VILA is a family of state-of-the-art vision language models (VLMs) for diverse multimodal AI tasks across the edge, data center, and cloud.

Repository: https://github.com/NVlabs/VILA
Canonical: https://ross.abutalabs.com/products/vila
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-03-12T01:04:00+00:00

## Health v2 (maintenance only)
Score: 57/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 71, release rhythm 35, longevity 65
- inputs: {"age_days": 922, "days_push": 175, "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 3857, forks 331 (observed 2026-08-28T04:08:27.448661+00:00)

## What it is
VILA is a family of open-source vision language models (VLMs) optimized for efficient video and multi-image understanding, spanning edge, data center, and cloud deployment. The repository provides training, fine-tuning, and inference code along with pretrained model weights in multiple sizes.

## Use cases
- run a vision language model for image question answering
- understand and summarize long videos with an LLM
- deploy a multimodal model on edge devices with 4-bit quantization
- fine-tune a VLM on custom image and video data
- compare multiple images in one conversation with an LLM
- build a medical or domain-specific multimodal assistant

## When to choose
- you need state-of-the-art open VLMs for video or multi-image understanding
- you want efficient deployment across edge to cloud with quantized checkpoints
- you need a codebase supporting training, RL fine-tuning, and long-context multimodal inference

## When to avoid
- you need a lightweight general-purpose LLM without vision capabilities
- you require commercially licensed model weights (models are CC BY-NC 4.0)
- you lack GPU hardware, as training and inference are GPU-intensive

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, llm-inference, llm-training, computer-vision, nlp
- domain: artificial-intelligence, large-language-models, computer-vision, deep-learning
- platform: python, cloud
- tags: vision-language-model, multimodal, video-understanding, multi-image, model-weights, quantization, awq, edge-deployment, video, linux, gpu

## Member repositories
- NVlabs/VILA (main) score 57

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:27.448661+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 2026-08-29T18:25:55.075292+00:00, confidence not recorded.
  - readme: https://github.com/NVlabs/VILA (fetched 2026-08-28T04:08:27.448661+00:00, sha c1bab0878051)
- Data as of 2026-08-30T08:39:29.467469+00:00.