Give your team AI superpowers, without handing over your data.

The software is free and open source. Build from source and pay Grengin nothing. Prefer one click? The AWS/Azure Marketplace image meters about $0.001 per vCPU-hour (about $3/month on a typical 4-vCPU VM), billed through your existing cloud account. Either way, you pay your cloud and your AI providers directly.

No seats, no minimums, no named plans. Whether you're 5 people or 500, the price doesn't change—cost scales with the compute and AI usage you actually consume. ChatGPT Enterprise pricing starts at ~150 seats (~$9,000/month) before anyone sends a single prompt.

Never, and no. You bring your own API keys (OpenAI, Anthropic, Google, and more) and pay each provider's published rates directly—Grengin adds zero margin. Optional test credits exist only as a convenience for trying a model you don't yet have a key for, billed at exact provider cost. They're never required, and your own keys keep working regardless of credit balance.

When you self-host (from source or the Marketplace image), Grengin runs as a single-tenant deployment inside your own AWS or Azure account. Conversations stay in your tenant, and with your own keys prompts travel straight from your deployment to your AI provider. Grengin Inc. is never in the data path, and never sees your prompts, completions, or logs.

No. Grengin's enterprise AI governance policy ensures your business data is never used to train AI models, and because you connect with your own API keys, your requests are covered by your direct agreements with each provider—agreements that exclude training on API traffic. The entire codebase is public on GitHub, so your security team can verify there's no hidden behavior.

Self-hosted Grengin runs inside your existing security and compliance boundary: your IAM, your encryption keys, your network controls, your cloud region. Need data residency? Deploy in whichever region your regulations require—the data never leaves your tenant. Every AI interaction lands in exportable audit trails, PII detection helps stop sensitive data from leaving, and open-source code means auditors can inspect exactly what's running.

OpenAI, Anthropic (Claude), Google Gemini, Mistral, Groq, and Cerebras: six model providers in one governed platform, switchable per task. Your team always uses the best model for the job, under one set of governance policies, with your own keys.

A full AI governance layer: SSO and centralized user management, policies by department or project, real-time cost tracking with budget alerts and hard caps per user or team, PII detection, and complete audit trails. AI for all, control for IT, insights for executives.

Jira, Confluence, and Google Drive out of the box—plus anything that speaks MCP, the open standard for connecting AI to your tools. SSO works with your existing identity provider. And yes, we're fully at home in a Google Workspace shop.

About five minutes, no DevOps team required. Launch the one-click Marketplace image in your AWS or Azure account, add your API keys and SSO, and invite your team. Engineers who want full control can build from source instead.

Yes—and there's nothing to commit to. Explore the interactive demo without signing up, or deploy the Marketplace image: it's metered by the hour with no contract, so you can shut it down anytime. The build-from-source path is free, forever.

No—by design. Grengin is open source AI Platform, your deployment runs in your own cloud, your conversations live in your own tenant, and AI access runs on your own provider keys. If you ever leave, you keep your data and your keys. Nothing breaks.

Community support is free: GitHub and the Grengin community. Optional paid support is available for any self-hosted deployment, purchased separately, so you only pay for it if you want it. No bundled tiers, no contract terms on the software itself.