VideoLLaMA3-7B

Name: VideoLLaMA3-7B
Author: DAMO-NLP-SG

by DAMO-NLP-SG

Video understanding foundation model with efficient long-video processing

2Kdl/month

77likes

8.0Bparams

HuggingFace Run it on your own data, free

Identifiers

Model ID

DAMO-NLP-SG/VideoLLaMA3-7B

Feature URI

mixpeek://video_extractor@v1/damo_videollama3_7b_v1

Overview

VideoLLaMA3 is a frontier multimodal model for image and video understanding from Alibaba DAMO Academy. It uses a vision-centric architecture with a 4-stage training pipeline including video-centric fine-tuning.

The model reduces vision tokens based on frame similarity for efficient long-video processing, making it practical for indexing hours of footage without proportional compute cost.

Architecture

7B parameter model with vision-centric design. 4-stage training: image pretraining → image SFT → video pretraining → video SFT. Adaptive token reduction based on inter-frame similarity for long videos.

Mixpeek SDK Integration

// No extractor parameter takes a Hugging Face model id (checked against
// GET /v1/discovery/extractors, which returns 13), so VideoLLaMA3-7B runs
// on your side and the output is upserted through POST
// /v1/namespaces/{namespace_id}/documents/upsert. On Enterprise the other
// path is to upload the weights instead: POST /v1/namespaces/{id}/models
// accepts the huggingface format and a custom plugin loads them.
const res = await fetch(
  "https://api.mixpeek.com/v1/namespaces/ns_your_namespace/documents/upsert",
  {
    method: "POST",
    headers: {
      Authorization: "Bearer API_KEY",
      "Content-Type": "application/json",
    },
    body: JSON.stringify({
      collection_id: "col_your_collection",
      documents: [
        {
          document_id: "asset-00412",
          // The model produces text, so it lands in payload. Give the
          // collection a text vector index and embed that text to make it
          // searchable rather than only filterable.
          payload: { extracted_text: modelOutput, source_key: "archive/2026/asset-00412" },
          vectors: { "multimodal-embedding": embeddingOfModelOutput },
        },
      ],
    }),
  },
);

// Managed alternative, if this exact model is not the requirement:
// universal_extractor@v1 runs google/gemini-embedding-2
// (3072-d) over a bucket, with no inference of your own.