#6464th of 67 in RAG, knowledge and search

videorag

Desktop app and algorithm for question answering over very long videos

Stars
3.5k
License
custom license
Last commit
Mar 2026
Language
Python
Stars, last 29 weeks+650
25 Apr 2026: 2,876 stars2 May 2026: 2,923 stars9 May 2026: 2,950 stars16 May 2026: 2,972 stars23 May 2026: 2,983 stars30 May 2026: 2,998 stars6 Jun 2026: 3,016 stars13 Jun 2026: 3,034 stars20 Jun 2026: 3,041 stars27 Jun 2026: 3,052 stars4 Jul 2026: 3,067 stars11 Jul 2026: 3,121 stars18 Jul 2026: 3,160 stars25 Jul 2026: 3,185 stars1 Aug 2026: 3,204 stars8 Aug 2026: 3,241 stars15 Aug 2026: 3,263 stars22 Aug 2026: 3,310 stars29 Aug 2026: 3,322 stars5 Sept 2026: 3,334 stars12 Sept 2026: 3,350 stars19 Sept 2026: 3,366 stars26 Sept 2026: 3,374 stars3 Oct 2026: 3,380 stars10 Oct 2026: 3,384 stars17 Oct 2026: 3,402 stars24 Oct 2026: 3,418 stars31 Oct 2026: 3,501 stars7 Nov 2026: 3,517 stars14 Nov 2026: 3,526 stars

2,876 on 25 Apr 20263,526 on 14 Nov 2026

From GitHub's daily star counts.

Checked on GitHub.

Overview

Vimo is an Electron desktop app backed by a Python VideoRAG server. It indexes videos into a multi-modal knowledge graph and answers natural-language questions about them. The VideoRAG algorithm (arXiv 2502.01549) builds on nano-graphrag, LightRAG and ImageBind, and the repo includes the LongerVideos benchmark and evaluation scripts. Prebuilt app releases are not available yet, so setup is from source.

Who it is for: Researchers and developers working on long-video question answering

Strengths

  • Targets videos of hundreds of hours on a single RTX 3090 (24GB)
  • Includes LongerVideos benchmark (164 videos, ~134.6 hours) with reproduction scripts
  • Reports 60.2% on Video-MME long track versus 51.8% to 56.3% for MiniCPM baselines
  • Supports queries across multiple videos

Weaknesses

  • No downloadable release yet; macOS Apple Silicon beta is only announced
  • Must run a Python backend and Electron frontend from source
  • License is not detected (NOASSERTION), so reuse terms are unclear
  • Main README gives no model provider list; setup details live in subfolders

What it needs

  • GPU required
  • Needs Python, Electron, Conda, ImageBind, nano-graphrag, LightRAG
  • Models: MiniCPM-o, MiniCPM-V, ImageBind

Also in RAG, knowledge and search

See all 67
Also in RAG, knowledge and search
RankProjectScore
1Open NotebookSelf-hosted NotebookLM alternative for research notebooks, chat, and podcasts42.1k starsMIT92 out of 100
2LightRAGGraph-based RAG framework with a REST server and WebUI40.8k starsMIT92 out of 100
3RAGFlowRAG engine with document parsing, agentic retrieval and citations Live demo ↗ (opens in a new tab)94.4k starsApache-2.090 out of 100
#4deeptutorTutoring workspace with knowledge bases, guided learning, memory and quizzes60.2k starsApache-2.090 out of 100
#5MaxKBEnterprise knowledge-base agent platform with RAG, workflows and MCP tools23.2k starsGPL-3.089 out of 100