Estúdio de Clone + UGC com IA

Suas fotos e sua voz viram vídeos falando, sem precisar se gravar toda vez. É uma versão em app do que a AvatarHype ensina em curso (análise completa em docs/engenharia-reversa-avatarhype.md).

Fluxo em /clone:

Fotos → retrato base (nano-banana-edit): de 1 a 6 fotos geram um retrato 9:16 fiel à sua identidade.
Voz (opcional) (minimax-voice-clone): amostra de áudio + consentimento explícito → sua voz clonada.
Roteiro → vídeo: cena opcional (nano-banana-edit) → fala com sua voz ou uma voz pronta (minimax-speech-2.6-hd) → foto falando (infinitetalk-image-to-video). As etapas encadeiam sozinhas, por webhook ou por polling.

Os agentes de roteiro (estrategista, roteirista, diretor de cenas) estão em prompts/.

Rodar localmente

cp .env.example .env    # DATABASE_URL, NEXTAUTH_SECRET, UGC_API_KEY (MuAPI)…
npm install
npx prisma db push
npm run dev             # http://localhost:3000/clone

Sem WEBHOOK_URL público, a página /clone consulta o status dos jobs a cada 8 s (POST /api/jobs/[requestId]/refresh). Você também pode entrar só com a sua chave MuAPI; nesse caso os créditos internos não são cobrados.

Custos internos por etapa: src/lib/config.js → clone.costs.

Baseado no Open-AI-UGC (MIT). README original abaixo.

Open AI UGC — Free Open-Source Alternative to Arcads & MakeUGC

The free, open-source alternative to Arcads and MakeUGC. Generate AI UGC video ads with realistic AI actors using 4+ state-of-the-art video models — no per-seat pricing, no locked actor library, no vendor watermark.

Community: Join Reddit & Discord for discussions and support

🤖 Automate UGC ad generations with AI coding agents: muapi-skills — a library of skills that let agents like Claude Code, Codex, and other coding assistants drive image/video models end-to-end (script → generate → review → re-render) directly from your terminal. Perfect for shipping 100 ad variants overnight without touching a UI.

Related projects

Open-source Node based workflow builder -> https://github.com/SamurAIGPT/Vibe-Workflow

Open-source AI Clipping — turn any long-form YouTube video into viral-ready vertical shorts -> https://github.com/SamurAIGPT/AI-Youtube-Shorts-Generator

Open-source AI Design Agent -> https://github.com/Anil-matcha/Open-AI-Design-Agent

Schedule your UGC content automatically with AI -> https://github.com/Anil-matcha/Free-AI-Social-Media-Scheduler

Learn to monetize AI-generated content like this -> https://github.com/Anil-matcha/ai-creator-academy

🎨 Explore 50+ more open-source AI apps →

🌐 Live Demo & Deploy

Live Demo: open-ai-ugc.vercel.app

Follow the creator for updates

Open AI UGC is a free, open-source AI UGC video ad studio that brings the Arcads / MakeUGC workflow to everyone — without the $110–$500+/month per-seat subscription. Paste a script, upload a reference face or product image, pick a model, and ship a scroll-stopping 9:16 vertical ad for TikTok, Reels, or Shorts in under a minute. Powered by Muapi.ai, it supports image-to-video and text-to-video generation across models like Veo 3.1, Seedance 2, Grok Video, and Happy Horse 1 — all from a sleek, modern interface you can self-host, rebrand, and customize end to end.

Why Open AI UGC instead of Arcads or MakeUGC?

Free & open-source — no $110–$500/mo subscription, no vendor lock-in
Bring your own actors — upload any face or product image instead of being locked to a vendor roster
Model choice — switch between Veo 3.1, Seedance 2, Grok Video, and Happy Horse 1 per render
No watermark, no branding — 100% your domain, your studio, your client deliverables
Self-hosted — your data and renders stay on your infra, full creative control
Stripe & credits built-in — clone, swap branding, and sell it as your own UGC SaaS
Extensible — add any MUAPI model in ~10 lines, modify the UI, build agencies on top of it

✨ Features

AI UGC Video Studio — Paste a script, optionally upload up to 7 reference images (faces, products, scenes), pick a model, and generate a 9:16 vertical UGC ad. Reference images inline with @image1, @image2, etc.
Multi-model support — 4 leading video models pre-wired: Veo 3.1 (Google), Seedance 2 (ByteDance), Grok Video (xAI), and Happy Horse 1. Switch per render.
Image-to-Video & Text-to-Video — Animate a product shot, an AI-generated actor, or generate purely from text — your call.
Per-model parameter control — Aspect ratio (9:16, 16:9, 1:1, etc.), duration (3–30s), resolution (480p → 4K), and mode (fun / normal / spicy for Grok) — UI adapts automatically to each model's capabilities.
Async job pipeline — MUAPI delivers finished renders via webhook; the dashboard polls and updates live without page reloads.
Creations dashboard — Every generation is persisted per user with prompt, settings, status, and final video URL.
Credits system — Each user starts with 10 free credits; 1 credit per generation; insufficient-credit gating built in.
Stripe billing — 3-tier pricing page (Free / Pro $19.99 / Elite $49.99) with Checkout + webhook credit top-up. Swap the tiers to whatever you want and ship.
Google sign-in — NextAuth.js with Prisma adapter, ready to go. Add more providers in one file.
Responsive UI — Tailwind v4 + Framer Motion, works on desktop and mobile, glassmorphism design.

🎬 UGC Video Generation — Dual Mode

The studio automatically adapts based on whether you upload reference images:

Mode	Trigger	Use case	Prompt
Text-to-Video	No image uploaded	Generate a UGC ad purely from a script	Required
Image-to-Video	Reference image(s) uploaded	Animate a specific actor face / product shot / scene	Optional

Supported Models

Model	Provider	Type	Aspect Ratios	Duration	Resolution	Notes
Veo 3.1	Google	I2V	16:9, 9:16	8s	720p / 1080p / 4K	High-fidelity, realistic motion — best for premium client ads
Seedance 2	ByteDance	I2V	21:9, 16:9, 4:3, 1:1, 3:4, 9:16	4–15s	—	Character-reference support, multi-shot generation
Grok Video	xAI	I2V	9:16, 16:9, 2:3, 3:2, 1:1	6–30s	480p / 720p	Modes: fun / normal / spicy — strong for high-CTR hooks
Happy Horse 1	—	I2V	16:9, 9:16, 1:1, 4:3, 3:4	3–15s	720p	Fast, lifelike animation — great for batch iteration

Adding a new MUAPI model takes ~10 lines in src/app/page.js (append to the MODELS array) and one line in src/app/api/generate/route.js (add to MODEL_ENDPOINTS).

🖼️ Multi-Image Reference Input

UGC ads frequently need multiple references — actor face + product + setting. Open AI UGC supports up to 7 reference images per generation:

Drag-and-drop or click to upload — images upload asynchronously with live status pills
Inline references in your script — @image1 holding the bottle, walking through @image2
Per-image previews with one-click removal
Auto-routing — passed as both image_url (first image) and images_list (full array) so every MUAPI model receives them in the format it expects

💳 Credits & Stripe Billing

Out-of-the-box billing flow you can sell as a SaaS:

Plan	Price	Credits	Target
Starter	Free	10	Trial users
Pro Studio	$19.99/mo	500	Solo creators / freelancers
Elite Creator	$49.99/mo	1,500	Agencies / power users

Plans, prices, and features are defined in one file (src/app/pricing/page.js) — change them to whatever your business model needs. Stripe Checkout handles the payment, the webhook (src/app/api/webhook/stripe/route.js) increments credits on success.

🚀 Quick Start

Prerequisites

Node.js (v18+)
A PostgreSQL database (free tiers on Neon, Supabase, or Railway)
A Muapi.ai access key — copy the generated key value, not the label
A Google OAuth client (Client ID + Secret)
A Stripe account (test mode is fine to start)

Setup

# Clone the repo
git clone https://github.com/Anil-matcha/Open-AI-UGC.git
cd Open-AI-UGC

# Install dependencies
npm install

# Copy env template and fill in the values
cp .env.example .env

# Push the Prisma schema to your database
npx prisma db push

# Start the dev server
npm run dev

Open http://localhost:3000, sign in with Google, and start generating. Every new user starts with 10 free credits.

Environment Variables

# Postgres (Neon / Supabase / Railway / self-hosted)
DATABASE_URL="postgresql://user:password@host:port/dbname"
DIRECT_URL="postgresql://user:password@host:port/dbname"

# NextAuth — generate with: openssl rand -base64 32
NEXTAUTH_URL="http://localhost:3000"
NEXTAUTH_SECRET="..."

# Google OAuth
GOOGLE_CLIENT_ID="..."
GOOGLE_CLIENT_SECRET="..."

# MUAPI — https://muapi.ai/access-keys?utm_source=github&utm_medium=readme&utm_campaign=open-ai-ugc
UGC_API_KEY="..."
WEBHOOK_URL="https://your-deployment-url.com"  # MUAPI calls this when a render finishes

# Stripe
STRIPE_SECRET_KEY="sk_test_..."
NEXT_PUBLIC_STRIPE_PUBLISHABLE_KEY="pk_test_..."
STRIPE_WEBHOOK_SECRET="whsec_..."

# UI
NEXT_PUBLIC_THEME=light

Production Build

npm run build
npm run start

⚡ Deployment: Vercel & Production

This architecture is engineered explicitly for Vercel serverless environments.

🚀 Launching on Vercel: Step-by-Step

Database Provisioning: Create a new Postgres database (via Supabase or Neon) and retrieve connection URLs.
Project Creation: Import your GitHub repository (https://github.com/Anil-matcha/Open-AI-UGC) into the Vercel dashboard.
Configure Environment Variables: Add all variables from .env.example in Vercel. Set WEBHOOK_URL to your Vercel deployment URL (e.g. https://open-ai-ugc.vercel.app).
Deploy: Build with Vercel. Next.js page generation will run Prisma client generation automatically via our script config.
Database Push: Run npx prisma db push to synchronize database models before launching.
Webhooks Setup:
- Configure a Stripe Webhook pointing to https://open-ai-ugc.vercel.app/api/webhook/stripe selecting checkout.session.completed.
- Configure the MuAPI Webhook pointing to https://open-ai-ugc.vercel.app/api/webhook/muapi.

🏗️ Architecture

Open-AI-UGC/
├── src/
│   ├── app/
│   │   ├── layout.js                       # Root layout (Tailwind, NextAuth provider, Navbar)
│   │   ├── page.js                         # Main UGC generator — model picker, prompt, image upload
│   │   ├── dashboard/page.js               # User's generation history
│   │   ├── pricing/page.js                 # Stripe-powered pricing page
│   │   └── api/
│   │       ├── generate/route.js           # POST — kick off a UGC video generation
│   │       ├── upload/route.js             # POST — image upload → hosted URL
│   │       ├── creations/route.js          # GET — list user's creations
│   │       ├── creations/[id]/route.js     # GET — single creation (polled by frontend)
│   │       ├── webhook/muapi/route.js      # POST — MUAPI render-complete callback
│   │       ├── webhook/stripe/route.js     # POST — Stripe payment callback → top up credits
│   │       ├── checkout/stripe/route.js    # POST — create Stripe Checkout session
│   │       └── auth/[...nextauth]/route.js # NextAuth handler
│   ├── components/
│   │   ├── saas/Navbar.jsx                 # Top nav with credits badge & profile menu
│   │   ├── saas/AuthButtons.jsx
│   │   └── ui/                             # Button, Card primitives
│   └── lib/
│       ├── auth.js                         # NextAuth config (Google + Prisma adapter)
│       └── prisma.js                       # Prisma client singleton
├── prisma/
│   └── schema.prisma                       # User, Creation, Account, Session, VerificationToken
└── package.json

🔌 API Integration

Open AI UGC communicates with Muapi.ai using a webhook pattern:

Submit — POST /api/v1/{model-endpoint} with prompt, images_list, model parameters, and webhook_url
Wait — MUAPI runs the render asynchronously (8s–2min depending on model and duration)
Callback — MUAPI calls your webhook_url with the finished video URL and request ID
Update — /api/webhook/muapi/route.js finds the matching Creation row by request_id and updates its status + URL
Poll — the frontend polls /api/creations/[id] every 3 seconds and switches from spinner to video player as soon as status flips

Authentication uses the x-api-key header. Credits are decremented on submit (so failed renders cost a credit — you can adjust this in src/app/api/generate/route.js if you'd rather refund on failure).

🛠️ Tech Stack

Next.js 16 — App Router, server components, React 19
React 19 — Studio UI
Tailwind CSS v4 — Utility-first styling
Framer Motion — Animations and transitions
Prisma 7 — Type-safe Postgres ORM
NextAuth.js — Google OAuth with Prisma adapter
Stripe — Checkout + webhooks for credit top-up
Muapi.ai — AI video model API gateway (Veo, Seedance, Grok, Happy Horse, and more)

🤔 How is this different from Arcads and MakeUGC?

Open AI UGC is a community-driven, open-source alternative that gives you the same Arcads/MakeUGC workflow without the closed ecosystem and per-seat pricing.

	Arcads / MakeUGC	Open AI UGC
Cost	$110–$500+/mo per seat	Free (open-source) — pay only for MUAPI API usage
Actor library	Locked to vendor's roster	Bring your own — any face or product image
Model choice	Single in-house model	Veo 3.1, Seedance 2, Grok Video, Happy Horse 1 (swap per render)
Hosted version	SaaS only	Self-host on Vercel / your own infra
White-label	Vendor branding / templates	100% your brand, your domain, your client deliverables
Custom pricing	Fixed tiers	Stripe + 3 tiers wired up, edit one file to change
Source code	Closed	MIT licensed, fully hackable
Data privacy	Cloud-only	Your renders and user data stay on your infra
Reseller-friendly	No	Yes — built to be cloned, rebranded, and sold

📄 License

MIT — fork it, rebrand it, ship it.

🙏 Credits

Built with Muapi.ai — the unified API for AI image and video generation models.

Looking for a free, open-source alternative to Arcads or MakeUGC? Open AI UGC is a self-hostable AI UGC video ad studio you can customize, rebrand, and resell. No content filters, no vendor lock-in, no per-seat pricing.

Name		Name	Last commit message	Last commit date
.github/workflows		.github/workflows
docs		docs
prisma		prisma
prompts		prompts
src		src
.env.example		.env.example
.gitignore		.gitignore
LICENSE		LICENSE
README.md		README.md
jsconfig.json		jsconfig.json
next.config.mjs		next.config.mjs
package-lock.json		package-lock.json
package.json		package.json
postcss.config.mjs		postcss.config.mjs
prisma.config.ts		prisma.config.ts
teste notion		teste notion

Navigation Menu

Repository navigation

About

Repository files navigation

Estúdio de Clone + UGC com IA

Rodar localmente

Open AI UGC — Free Open-Source Alternative to Arcads & MakeUGC

Related projects

🌐 Live Demo & Deploy

✨ Features

🎬 UGC Video Generation — Dual Mode

Supported Models

🖼️ Multi-Image Reference Input

💳 Credits & Stripe Billing

🚀 Quick Start

Prerequisites

Setup

Environment Variables

Production Build

⚡ Deployment: Vercel & Production

🚀 Launching on Vercel: Step-by-Step

🏗️ Architecture

🔌 API Integration

🛠️ Tech Stack

🤔 How is this different from Arcads and MakeUGC?

📄 License

🙏 Credits

About

Releases

Packages

Contributors

Languages

Latest commit

History

Folders and files

Repository files navigation

Estúdio de Clone + UGC com IA

Rodar localmente

Open AI UGC — Free Open-Source Alternative to Arcads & MakeUGC

Related projects

🌐 Live Demo & Deploy

✨ Features

🎬 UGC Video Generation — Dual Mode

Supported Models

🖼️ Multi-Image Reference Input

💳 Credits & Stripe Billing

🚀 Quick Start

Prerequisites

Setup

Environment Variables

Production Build

⚡ Deployment: Vercel & Production

🚀 Launching on Vercel: Step-by-Step

🏗️ Architecture

🔌 API Integration

🛠️ Tech Stack

🤔 How is this different from Arcads and MakeUGC?

📄 License

🙏 Credits

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages