Show HN: ChatRAG – Next.js and AI SDK starter to ship RAG chatbots faster (www.chatrag.ai)

ChatRAG is a Next.js starter kit that packages a complete retrieval-augmented generation (RAG) chatbot stack into a ready-to-deploy boilerplate — pay once and self-host unlimited chatbots. The dashboard accepts PDFs, Word, Excel and other files and automatically parses, chunks (via LlamaCloud), embeds (OpenAI embeddings) and indexes content into a Supabase HNSW vector store for ultra-fast semantic retrieval. The product includes built-in auth, persistent chat history, an embeddable web widget, WhatsApp connectivity (no Business account required), Stripe/Polar monetization, and UI customization without code. It also bundles extras for production chatbots: image/video/3D generation, web search + RAG, executable code Artifacts, OpenAI/STT and ElevenLabs TTS, MCP integrations (Zapier, n8n), and a localized UI in 14+ languages. For AI/ML teams and builders, ChatRAG lowers the barrier from months of engineering to minutes of setup, giving a full RAG pipeline with model-agnostic connectivity (OpenRouter support for 200+ models) and continual updates to state-of-the-art RAG techniques. Technically notable choices — LlamaCloud for intelligent chunking, OpenAI embeddings for precise semantics, and HNSW indexing in Supabase — mean faster, more contextual retrieval and easier scaling. The one-time, self-hosted model emphasizes data ownership and avoids vendor lock-in, making it appealing for startups, agencies, and enterprises that want rapid deployment, monetization, and full control over their knowledge bases.