# HKUDS/ViMax

"ViMax: Agentic Video Generation (Director, Screenwriter, Producer, and Video Generator All-in-One)"

Repository: https://github.com/HKUDS/ViMax
Canonical: https://ross.abutalabs.com/products/vimax
Homepage: https://arxiv.org/abs/2606.07649
Language: Python
License: MIT
License Family: permissive
Topics: agentic-aigc, video-generation
Last push: 2026-07-29T08:56:47+00:00

## Health v2 (maintenance only)
Score: 72/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 95, release rhythm 61, longevity 37
- inputs: {"age_days": 521, "days_push": 35, "days_rel": 44, "gap_med": null, "n_releases_24m": 1}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 12108, forks 1823 (observed 2026-08-28T04:10:51.759341+00:00)

## What it is
ViMax is an agentic video generation framework that uses coordinated multi-agent collaboration (director, screenwriter, producer, video generator roles) to produce long-form narrative videos. It combines a hierarchical narrative engine with retrieval-augmented generation and dependency-aware visual consistency tracking to maintain story coherence and character/scene continuity across multi-scene timelines.

## Use cases
- generate long-form AI videos with consistent characters across scenes
- automate video storytelling with AI agents acting as director and screenwriter
- create multi-scene narrative videos instead of short clips
- maintain visual consistency in AI-generated video sequences
- produce scripted videos with audio and narrative structure automatically
- research agentic AIGC pipelines for video production

## When to choose
- you need extended narrative videos with story structure, not just seconds-long clips
- character and environmental consistency across scenes is critical
- you want an end-to-end pipeline covering scripting, planning, and video generation
- you're researching multi-agent collaboration for generative video

## When to avoid
- you only need quick single short clips from a text prompt
- you lack GPU resources or access to video generation models
- you need real-time or live video generation
- you want a polished GUI video editor rather than a Python framework

## Facets
- artifact type: framework
- maturity: active
- function: video-processing, agent-framework, rag, machine-learning, llm-inference
- domain: artificial-intelligence, deep-learning
- platform: python, cross-platform
- tags: agentic-aigc, video-generation, multi-agent, narrative-planning, visual-consistency, long-form-video, storytelling, ai-agents, video, retrieval-augmented-generation, gpu

## Member repositories
- HKUDS/ViMax (main) score 72

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:51.759341+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 2026-08-29T17:14:55.861206+00:00, confidence not recorded.
  - readme: https://github.com/HKUDS/ViMax (fetched 2026-08-28T04:10:51.759341+00:00, sha 2718e4387d0c)
  - homepage: https://arxiv.org/abs/2606.07649 (fetched 2026-08-29T08:12:15.356211+00:00, sha d02539a14524)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T08:12:15.365628+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T08:12:15.369212+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T08:12:15.371119+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T08:12:15.367613+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.