whatisgithub

What is movie-gen?

dawndrain/movie-gen — explained in plain English

Analysis updated 2026-05-18

29PythonAudience · vibe coderComplexity · 3/5Setup · moderate

In one sentence

A pipeline that uses several AI services and Claude Code to generate short films with consistent characters, voices, and scenes.

Mindmap

mindmap
  root((movie-gen))
    What it does
      AI short films
      Consistent characters
      Claude Code driven
    Tech stack
      Python
      Higgsfield
      ElevenLabs
      ffmpeg
    Use cases
      Short film generation
      Animatic previews
      Template for new films
    Audience
      Vibe coders
      Indie filmmakers

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

What do people build with it?

USE CASE 1

Generate a short AI film with consistent characters and voices across many shots.

USE CASE 2

Preview an entire film cheaply as an animatic before generating expensive video clips.

USE CASE 3

Let Claude Code drive the whole filmmaking pipeline from script to assembled video.

USE CASE 4

Reuse the worked example project as a template for a new film idea.

What is it built with?

PythonffmpegHiggsfieldElevenLabsGemini

How does it compare?

dawndrain/movie-genadityasharmadotai-hash/docs-reader-rag-agentalekseiul/hermes-researcher-agent
Stars292929
LanguagePythonPythonPython
Setup difficultymoderateeasymoderate
Complexity3/52/52/5
Audiencevibe codervibe coderresearcher

Figures from each repo's GitHub metadata at analysis time.

How do you get it running?

Difficulty · moderate Time to first run · 1h+

Requires paid Higgsfield and ElevenLabs accounts with credits, Gemini API key is optional for quality checks.

No license information was found in the README.

So what is it?

movie-gen is a working pipeline for making short AI generated films, roughly five to twelve minutes long with dozens of shots, while keeping the same characters and voices consistent across the whole film. It stitches together several AI services: Seedance for video, Nano Banana Pro for images, Sonilo for music, all accessed through a tool called Higgsfield, plus ElevenLabs for character voices, Gemini for automatically checking audio and image quality, and the video tool ffmpeg to assemble everything into a finished file. The process moves through clear stages. First you generate character portraits to keep faces consistent, then a starting image for each scene using those portraits as a reference, then cast and record voices for the script. Before spending money on full video generation, the whole film is put together cheaply as an animatic, meaning still images paired with the recorded voice lines, so you can catch story and pacing problems early. Only after that do the expensive video clips get generated, followed by adding music and sound, assembling everything with ffmpeg, and reviewing a rough cut to decide what still needs fixing. About ten films have already been made this way, including an eleven minute comedy called The Long Game, which is included in the repository as a full worked example with its scripts and iteration history. The project is explicitly designed to be driven by Claude Code rather than run by hand. A person mainly needs accounts with Higgsfield and ElevenLabs, then Claude Code reads the included playbook document, asks what kind of movie is wanted, installs tools, and drives most of the pipeline, leaving the human to review previews and give notes.

Copy-paste prompts

Prompt 1
Walk me through the movie-gen pipeline from character portraits to a finished film.
Prompt 2
Help me set up Higgsfield and ElevenLabs accounts to start a new movie-gen project.
Prompt 3
Explain how the animatic stage in movie-gen saves money before full video generation.
Prompt 4
Show me how to use the worked example in long_game as a template for my own film.

Frequently asked questions

What is movie-gen?

A pipeline that uses several AI services and Claude Code to generate short films with consistent characters, voices, and scenes.

What language is movie-gen written in?

Mainly Python. The stack also includes Python, ffmpeg, Higgsfield.

What license does movie-gen use?

No license information was found in the README.

How hard is movie-gen to set up?

Setup difficulty is rated moderate, with roughly 1h+ to a first successful run.

Who is movie-gen for?

Mainly vibe coder.

Open on GitHub → Ask about another repo

This repo across BitVibe Labs

Verify against the repo before relying on details.