NextGen Relay AI
AI SaaS Platform — 125+ Models, One API. Chat, image, video, audio, and music generation through one account, one credit system, and one OpenAI-compatible API key.
nextgenrelay.com
Watch the Platform Demo
A guided tour of NextGen Relay AI — chat, image, video, audio, and music generation, all from one account with one credit system.
The Challenge
AI models are fragmented across dozens of providers — OpenAI, Google, Anthropic, Midjourney, ElevenLabs, Suno, and more. Developers need separate API keys, billing accounts, and integration code for each one. Non-technical users can't access most of them at all.
The goal: Build a unified platform where anyone can access 125+ AI models — chat, image, video, audio, music — through one account, one credit system, and one API key.
Why It Matters
One Account, All AI
No more juggling 15 API keys, 15 billing accounts, and 15 different dashboards. Your users get everything in one place.
Built-In Monetization
Credit-based billing, subscription plans, and pay-as-you-go top-ups — all wired up from day one. Start generating revenue immediately.
Production-Grade Security
GDPR consent, rate limiting, webhook verification, RLS policies — the compliance layer that lets you ship to enterprise clients.
What I Built
Eleven feature modules — from multi-model generation to SEO, GDPR compliance, and content marketing.
Multi-Model Chat
- 125+ AI models (GPT-4o, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, Qwen) in one interface
- Model selector with provider grouping and tier badges (flagship / standard / fast / budget)
- Real-time streaming responses
- Conversation history with export
Image Generation
- Text-to-image: GPT Image 2, DALL-E 3, Flux, Midjourney, Nano Banana, Seedream, Hidream
- Image-to-image: upload a reference and transform with AI
- Smart Retouch: AI-powered image enhancement and editing
- Clothing Studio: virtual try-on with AI garment fitting
- Quality and size controls per model
Video Generation
- Models: Seedance 2.0, Veo 3.1, Sora 2, Kling, Luma, Higgsfield, Runway, Pixverse, Vidu
- Duration and resolution controls per model
- Async task polling with progress tracking
- Preview and download when complete
Audio Suite
- Text-to-Speech: ElevenLabs, MiniMax, Azure Neural TTS, GLM TTS — 724+ voices, 100+ languages
- Speech-to-Text: Whisper, GPT-4o Transcribe, Mini Transcribe, speaker diarization
- Music Generation: Suno V5.5 with custom mode (lyrics + style tags)
- Audio playback with waveform visualization
Podcast Creator
- AI-powered podcast generation from topic prompts
- Multi-speaker conversation generation
- Background music integration
- Export to MP3
Visual Bot Builder
- No-code bot creation with custom system prompts
- Select any AI model as the bot's brain
- Custom avatar and personality
- Shareable bot links
OpenAI-Compatible API
- Developers get a single API key
- Access all 125+ models via the standard OpenAI SDK format
- Same credit system as the web interface
- Full API documentation with examples
Credit-Based Billing
- 1 USD = 100 credits, pay-as-you-go top-ups ($10 / $25 / $50 / $100 + custom)
- Subscriptions: Pro ($29/mo → 3,190 credits), Business ($99/mo → 10,890 credits)
- Per-model cost calculation (each model has a different credit cost)
- Top-up credits never expire; subscription credits reset monthly
SEO & Discoverability
- JSON-LD structured data (SoftwareApplication schema) for Google rich results
- Dynamic XML sitemap with per-page priorities
- Smart robots.txt — allow public pages, disallow authenticated routes (/api, /chat, /billing)
- Open Graph + Twitter Card metadata (1200x630) for social sharing
- Canonical URLs, 12 targeted keywords, and Bing webmaster verification
Security & Compliance
- GDPR cookie consent with GA4 Consent Mode v2 (privacy-first analytics)
- Custom rate limiting with 4 presets — AI generation (30/min), uploads (10/min), auth (5/min), API (60/min)
- Per-user rate limiting (API key > user ID > client IP priority)
- Rate limit headers (X-RateLimit-Limit / Remaining / Reset, Retry-After)
- OpenAI-compatible 429 error responses for API users
Content Marketing
- MDX blog system with 3 SEO-optimized articles (intro, credits guide, image model guide)
- Full API documentation section
- Model library page with provider/category filtering
- Legal pages: Privacy Policy, Terms of Service, Cookie Policy, Refund Policy
- Dark/light theme switching with system preference detection
Inside the Platform
Multi-Model Chat
Model selector with provider grouping, tier badges, and real-time streaming.
Image Generation
Text-to-image and image-to-image across 7+ image models with per-model controls.
Video Generation
Seedance, Veo, Sora, Kling and more with async task polling and progress tracking.
Audio Suite
TTS/STT tabs plus Suno music generation with waveform playback.
Settings & Billing
Credit packs, subscription plans, and real-time balance tracking.
Landing & Pricing
Pay-as-you-go top-ups and Pro / Business subscription tiers.
Now on Android
The full Relay AI experience, rebuilt as a native Android app. All 8 AI creative tools — chat, image, video, audio, retouch, try-on, and podcast — powered by 125+ models, in your pocket.
A real native app, not a webview shortcut
Every screen, straight from the app
Real screenshots from the Android build — no mockups.
Tech Stack
Architecture Highlights
The systems that keep billing accurate, emails reliable, and 125+ models routed correctly.
Credit Billing System
- Real-time deduction per model, each with its own credit cost
- Subscription credits reset monthly via pg_cron (automated Supabase scheduler)
- Top-up credits never expire — a separate wallet from subscriptions
- PayPal webhook signature verification for payment security
- Atomic credit deduction with database-level locking
Email Automation (n8n)
- 6 transactional email types: top-up, subscription, renewal, payment failed, cancelled, low-credit warning
- Welcome email for new users on first signup
- Real-time admin payment notifications
- All routed through one n8n workflow with event-based switching
- Gmail priority headers for inbox placement
AI Model Routing
- Proprietary gateway mapping model IDs across multiple AI providers
- Single OpenAI-compatible endpoint for developer access
- Async polling for long-running tasks (video, music generation)
- Per-model credit cost calculation and real-time deduction
- Error handling with automatic retry and fallback
Authentication
- Supabase Auth with PKCE flow
- Email/password + Google OAuth + Apple OAuth
- Secure session storage (expo-secure-store on mobile)
- Protected routes with middleware
Security
- PayPal webhook signature verification
- Supabase Row Level Security (RLS) on all tables
- Rate limiting on API endpoints
- GDPR-compliant cookie consent with GA4 Consent Mode v2
- Atomic database operations for billing integrity
The Results
“Before NextGen Relay, integrating AI meant managing dozens of API keys and billing accounts. Now everything — chat, image, video, audio — lives under one roof with one credit system. It's the platform I wish existed when I started building with AI.”— John Michael De La Pena, Creator of NextGen Relay AI
Want a similar AI-powered platform?
Let's talk about building a unified, credit-based AI product for your business.