Grengin

🚀 Grengin: A Self-Hosted, Multi-LLM AI Workspace for Governed Teams Grengin is an open-source, self-hosted AI workspace that gives organizations governed access to multiple large language model providers — including OpenAI, Anthropic Claude, Google Gemini, and Mistral — through a single, branded chat interface. Instead of routing sensitive data through third-party SaaS platforms, Grengin runs entirely inside a company's own infrastructure, with traffic flowing directly from that environment to the model providers of choice. IT teams keep full visibility and control, while end users get a unified, familiar chat experience no matter which model is powering a given conversation. The platform closes a common gap: most organizations either lock employees out of AI tools over data-governance concerns, or let AI usage sprawl uncontrolled across dozens of personal accounts and browser extensions. Grengin is the middle path 🎯 — enterprise-grade governance without enterprise pricing, seat minimums, or a heavy DevOps lift. ⚙️ CORE FEATURES 🔀 Multi-provider LLM routing — swap or combine OpenAI, Anthropic, Google, Mistral, and open-source models without changing the frontend 🔐 Governed access — SSO (Google, Azure AD), role-based permissions, and department- or project-level policies 🛡️ Security & compliance — built-in PII detection and tamper-evident audit logs 💰 Cost control — real-time usage tracking and budget caps by user, department, or project 🔗 Tool integrations — connects to existing workflows via MCP (Model Context Protocol), Jira, Confluence, Slack, and Google Drive 🔍 Semantic search across conversation history, plus a personal prompt library for end users 📊 Admin dashboard for user management, provider configuration, and budgets 🔑 Bring-your-own API keys on every path, so Grengin itself never sits in the data path when self-hosted 🛠️ WAYS TO DEPLOY GRENGIN Grengin can typically be live in about five minutes, with several deployment paths depending on how much infrastructure control an organization wants: 1️⃣ Cloud marketplace image (one-click) Launch a pre-built image from AWS Marketplace or Azure Marketplace. PostgreSQL, Nginx, Certbot, and Grengin arrive pre-installed and pre-configured, with OAuth/SSO wiring already in place. On first boot, a setup wizard walks through domain, SSL, OAuth, and AI provider configuration, and the app goes live at the same URL. Every instance also qualifies for a free *.ongrengin.com subdomain, claimable during setup. 2️⃣ Grengin Cloud (managed) For teams that would rather not manage servers, patching, or uptime themselves, Grengin Cloud offers a fully managed version of the platform while preserving the bring-your-own-API-key model. 3️⃣ Run from source (free, fully self-managed) Organizations wanting maximum control can clone the open-source repository and run Grengin entirely on their own infrastructure at no cost, with full ability to audit, fork, or modify the codebase. Across every deployment option, admins keep the same controls: which AI providers and models are enabled, who can access them, how much each department can spend, and which MCP tool servers are available. 👥 WHO IT'S FOR Grengin is built for small and mid-sized organizations, as well as larger enterprises, nonprofits, government agencies, and freelancers, who want governed, auditable AI access without committing to a heavyweight enterprise AI suite. It suits IT and finance teams that need budget and compliance controls 📈, and end users who just want a clean, fast, multi-model chat interface with tool integrations and file uploads. Pricing starts at $0 for the self-hosted, open-source path, with paid options for teams that prefer a managed cloud deployment. ✨ WHY IT'S DIFFERENT Most AI governance tools either force a single vendor or require enterprise contracts to get basic controls like SSO, budgets, and audit logs. Grengin's self-hosted-first approach means the data never has to leave an organization's own environment unless the org chooses a managed path — and even then, API keys stay in the customer's hands 🔒. That mix of openness, fast setup, and enterprise controls at startup-level simplicity is the core of its pitch.