OpenReels

Open-source AI pipeline that turns any topic into a publish-ready YouTube/Instagram/TikTok Short. One command. Research, script, voiceover, AI-generated visuals, AI-generated music, animated captions, scene transitions. Out comes a vertical MP4, ready to upload.

Web UI with live pipeline visualization, REST API, Docker, and CLI. Bring your own API keys.

Demo

openreels "the apollo 13 disaster, from explosion to miraculous return" --provider google

Web UI

Live pipeline visualization. Watch research, script, voiceover, visuals, music, and assembly stages stream in real time.

Screenshots

Topic input with archetype gallery	Live pipeline with storyboard
Completed: video player, cost breakdown, quality review	Gallery with multiple generations

How it works

Give it a topic. It handles everything:

Every stage streams live progress to the web UI. You watch the AI research, write, paint, compose, and render in real time.

Provider flexibility

Mix and match providers or go all-in on one ecosystem:

One key, everything Google: --provider google sets LLM, images, TTS, video, and music to Google APIs with a single GOOGLE_API_KEY.

Zero-cost voiceover: --provider local uses Kokoro for free local TTS. No API key needed.

Quickstart

Web UI (recommended)

git clone https://github.com/tsensei/OpenReels.git
cd OpenReels
cp .env.example .env   # fill in your API keys
docker compose up      # starts Redis + API + Worker
# Open http://localhost:3000

Type a topic, pick your providers, and watch the pipeline run. Research, script, voiceover, visuals, music, and assembly stages stream live to the browser. Download the final video when it's done.

Docker CLI (single run)

cp .env.example .env   # fill in your API keys
docker run --env-file .env --shm-size=2gb -v ./output:/output ghcr.io/tsensei/openreels "5 stoic lessons that changed my life"

Or run through Docker Compose:

docker compose run worker npx tsx src/index.ts --yes "5 stoic lessons that changed my life"

Local development

Prerequisites: Node.js 22+, pnpm, ffprobe

git clone https://github.com/tsensei/OpenReels.git
cd OpenReels
pnpm install
cp .env.example .env   # fill in your API keys

# Full pipeline with AI music
pnpm start "the fall of the Roman Empire" --provider google
Free local TTS, no API spend on voiceover
pnpm start "5 stoic lessons" --provider local
Dry run (outputs DirectorScore JSON, no asset generation)
pnpm start "your topic" --dry-run
Specific archetype and provider combo
pnpm start "your topic" --archetype anime_illustration --provider openai
Creative direction file (guide the AI with a brief)
pnpm start "deep sea exploration" --direction examples/direction-brief.md
Replay from a previous score (skip research + director, re-render)
pnpm start "your topic" --score output/2026-04-10-111939-.../score.json

API keys

Minimum to run (pick one LLM + one TTS):

• ANTHROPIC_API_KEY or OPENAI_API_KEY or GOOGLE_API_KEY — Anthropic / OpenAI / Google AI Studio
• ELEVENLABS_API_KEY or INWORLD_TTS_API_KEY — ElevenLabs / Inworld. Or use --tts-provider kokoro (free, no key), --tts-provider openai-tts, or --tts-provider gemini-tts
• GOOGLE_API_KEY — also needed for Gemini image generation, AI video (Veo), AI music (Lyria), and Gemini TTS

Optional: PEXELS_API_KEY (Pexels), PIXABAY_API_KEY (Pixabay) for stock footage, FAL_API_KEY (fal.ai) for Kling video generation

CLI flags

Cost transparency

Before spending any money, the pipeline shows a detailed cost breakdown and asks for confirmation:

Estimated cost: $0.686
  LLM:    $0.0029 (7 calls)
  TTS:    $0.0171 (853 chars)
  Images: $0.3030 (3 AI images @ $0.101/ea)
  Video:  $0.3000 (1 AI videos)
  Music:  $0.0802 (Lyria AI generation)
  Stock:  free

After rendering, actual cost is computed from real token usage. Use --dry-run to preview the DirectorScore without spending anything.

Archetypes

14 visual styles that control colors, captions, motion, lighting, and AI image prompting. Same topic, four different archetypes:

Cinematic Documentary	Anime Illustration	Surreal Dreamscape	Vintage Snapshot

Background

OpenReels is a full rewrite of ReelMistri, a CLI pipeline originally built for Bangla-language YouTube Shorts automation. ReelMistri proved the concept: one command, fully produced video, language-aware scripts, culturally coherent visuals, proper Bengali text rendering.

The rewrite moves from Python to TypeScript for native Remotion integration. Cleaner video rendering, better developer experience, no Python-to-TypeScript bridge.

Status

v0.17.0 shipped. See CHANGELOG.md for full version history and TODOS.md for known issues and roadmap.

Star History

If OpenReels is useful to you, consider giving it a star. It helps others discover the project.

License

This project is licensed under the MIT License.