Skip to content

yhinai/flowMotion

Repository files navigation

FlowMotion

AI-powered video generator — from prompt to MP4 in minutes

Deploy to DO


What is FlowMotion?

FlowMotion is an autonomous AI video production pipeline. Describe a video in plain language and the system writes the script, generates cinematic clips, adds narration, and renders a final MP4 — all without manual editing.

Three paths to create video:

Path Description
AI Video (Veo 3) Full autonomous pipeline: Gemini writes the script, Veo generates clips, ElevenLabs narrates, Remotion renders
Remotion-Only Bring your own clips — Remotion composes them with text overlays and transitions
Upload & Edit Upload existing video and use the AI Director chat to request edits

Sponsor Integrations

Sponsor Integration
DigitalOcean App Platform deployment (.do/app.yaml) + Spaces CDN storage for all video assets
assistant-ui AI Director chat interface (@assistant-ui/react) + generation activity heatmap (heat-graph)
Nexla Live data pipelines feed trending topics (news, crypto, weather) for autonomous video generation
Augment Code AI-assisted development tooling used throughout the build

Architecture

See docs/flowmotion_architecture_v4.mermaid for the full system diagram.

Core flow: User prompt → Gemini script → Veo clips + ElevenLabs narration → Remotion composition → MP4 output


Quick Start

git clone https://github.com/yhinai/flowMotion.git
cd flowMotion
cp .env.example .env
# Fill in your API keys (see table below)
npm install
npm run dev

Open http://localhost:3000 to access the app.


Environment Variables

Variable Purpose Where to get it
GEMINI_API_KEY Gemini + Veo video generation Google AI Studio
ELEVENLABS_API_KEY TTS narration and sound effects ElevenLabs Settings
GOOGLE_CLOUD_PROJECT_ID Vertex AI / Lyria music generation GCP Console
GOOGLE_CLOUD_LOCATION Vertex AI region (default: us-central1) Same as above
DO_SPACES_KEY DigitalOcean Spaces access key DO API Settings
DO_SPACES_SECRET DigitalOcean Spaces secret key Same as above
DO_SPACES_BUCKET Spaces bucket name (default: flowmotion-videos) DO Spaces
DO_SPACES_REGION Spaces region (default: nyc3) Same as above
NEXT_PUBLIC_SUPABASE_URL Supabase project URL (fallback storage) Supabase Dashboard
SUPABASE_SERVICE_ROLE_KEY Supabase service role key Same as above
NEXLA_API_KEY Nexla live data pipeline access Nexla
NEXLA_NEWS_NEXSET_ID Nexla news flow ID Same as above
NEXLA_CRYPTO_NEXSET_ID Nexla crypto flow ID Same as above
NEXLA_WEATHER_NEXSET_ID Nexla weather flow ID Same as above
TELEGRAM_BOT_TOKEN Telegram bot interface (optional) @BotFather
REDIS_URL Redis for BullMQ job queue (optional) Your Redis instance
ENABLE_BULLMQ Enable persistent job queue (true/false) Set to true to enable
NEXT_PUBLIC_APP_URL Public app URL Your deployment URL

Tech Stack

Component Technology
Framework Next.js 15 (App Router)
Script Generation Google Gemini 2.5 Flash
Video Generation Google Veo 3
Narration ElevenLabs
Music Google Lyria 2
Video Composition Remotion
Chat UI assistant-ui
Live Data Nexla
Storage DigitalOcean Spaces / Supabase (fallback)
Bot Interface Telegram

Deploy to DigitalOcean

An App Platform spec is included at .do/app.yaml. Import the repo via the DigitalOcean App Platform and fill in the environment variables.


License

MIT

FlowMotion — Built at the Zero to Agent Hackathon, San Francisco, March 2026.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors