# DAMO-NLP-SG/VideoLLaMA2

VideoLLaMA 2: Advancing Spatial-Temporal Modeling and Audio Understanding in Video-LLMs

Repository: https://github.com/DAMO-NLP-SG/VideoLLaMA2
Canonical: https://ross.abutalabs.com/products/videollama2
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2025-01-23T04:20:22+00:00

## Health v2 (maintenance only)
Score: 25/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 3, release rhythm 35, longevity 58
- inputs: {"age_days": 824, "days_push": 587, "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 1307, forks 89 (observed 2026-08-28T04:04:19.013525+00:00)

## What it is
VideoLLaMA 2 is an open-source video large language model that adds spatial-temporal modeling and audio understanding to video-LLMs. It provides pretrained checkpoints, training and inference code, and demos for video question answering and captioning.

## Use cases
- answer questions about videos with an LLM
- generate captions for videos automatically
- build a multimodal chatbot that understands video and audio
- fine-tune a video-language model on custom data
- benchmark zero-shot video question answering
- analyze spatial-temporal content in video clips

## When to choose
- you need a ready-to-use open video-LLM with audio support
- you want strong zero-shot video QA and captioning baselines
- you need Apache-2.0 licensed multimodal model code and checkpoints

## When to avoid
- you only need image-only vision-language understanding
- you lack GPU resources for large multimodal models
- you need a lightweight production video analytics service rather than a research model

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, llm-inference, llm-training, video-processing, audio-processing, nlp
- domain: artificial-intelligence, large-language-models, computer-vision
- platform: python
- tags: video-llm, multimodal, video-question-answering, video-captioning, audio-understanding, spatial-temporal-modeling, research, video, audio, natural-language-processing, gpu, linux

## Member repositories
- DAMO-NLP-SG/VideoLLaMA2 (main) score 25

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:19.013525+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 2026-08-30T04:50:07.351768+00:00, confidence not recorded.
  - readme: https://github.com/DAMO-NLP-SG/VideoLLaMA2 (fetched 2026-08-28T04:04:19.013525+00:00, sha c605d550aae5)
- Data as of 2026-08-30T08:39:29.467469+00:00.