Skip to content

wangzai-double-milk /Vibefilming

A finished short film from one plain sentence, with AI review and reshoots
Video frameworkMITChinese README
Stars
163
30 days
New
Last push
2 months ago

About Vibefilming

One plain sentence is all it takes to run the whole short-film pipeline. VibeFilming breaks the brief into shots, builds character references, and generates each shot. An AI review checks every image and clip, then reshoots what fails. You'll need a paid Volcano Ark API key, and the docs and UI are Chinese-only.

Best for: Plain-brief creators who want a finished short film from one sentence, without writing prompts or editing by hand.

What it does

  • Runs the full pipeline: storyboard, shoot, review, retake, stitch, score
  • Shows a local Studio with chat on the left and a live workflow canvas on the right
  • Reviews every image, clip and cut, then backfills missing assets and reshoots
  • Builds multi-view reference packs and blocking diagrams to keep characters consistent
  • Generates shots in parallel when they don't need to follow the previous one
  • Adds BGM at the end through Volcengine BigMusic (optional credentials)

Quickstart

Rewritten from the README. Check the repo for the latest steps.

  1. 1
    Clone the repo and run bash setup.sh to build the venv, install dependencies and copy the config
  2. 2
    Open vibefilming.config.json and fill in the Volcano Ark ark.api_key
  3. 3
    Run source .venv/bin/activate, then python3 agentmain.py
  4. 4
    Open the Studio link it prints (default http://127.0.0.1:8764/) or the canvas at http://127.0.0.1:8765/canvas.html
  5. 5
    Type a brief like a 30-second pet-themed short, then exit with Ctrl+C or /exit

Models and languages

Models it works with

  • DeepSeek V4 Pro
  • Seed-Pro VLM
  • Seedream
  • Seedance 2.0
  • BigMusic

Interface and docs

  • Chinese

Alternatives

Other video framework projects people compare with Vibefilming.

calesthio

OpenMontage

Video studio that your AI coding assistant runs

Video framework
  • Google Veo
  • Kling
  • Seedance 2.0
  • +7
62k GitHub stars25 days agoAGPL-3.0

Editable video plans from a plain-language brief, rendered by your coding agent

Video framework
  • Seedance
  • Kling v2.1
  • Atlas Cloud
  • +3
497 GitHub stars9 days agoMIT

HKUDS

ViMax

Multi-shot AI video from an idea, a script, or a novel chapter

Video framework
  • Seedance 2.0 Fast
  • GPT Image 2
  • Google Omni
  • +4
13k GitHub starsyesterdayMIT

Writes a narrated short video from a topic: script, voiceover, footage, subtitles

Video framework
  • MiniMax H3
  • Seedance
  • Wan
  • +7
128k GitHub starstodayMIT

Featured on OpenMicroDrama

Do you maintain Vibefilming? Add this badge to your README so English speakers can find our write-up.

Featured on OpenMicroDrama
markdown
[![Featured on OpenMicroDrama](https://openmicrodrama.com/badges/featured.svg)](https://openmicrodrama.com/projects/wangzai-double-milk-vibefilming)

The badge links to this page. Want your description changed? Email support@openmicrodrama.com.

Reviewed Oct 1, 2026. We write these descriptions ourselves. The repo's own docs are the final word.