TSLHostedAI

Run a local AI platform that feels like a hosted API.

This local stack uses Next.js, Prisma, PostgreSQL, Auth.js, and Ollama to expose a multi-user AI gateway with API keys, quotas, a user portal, an admin portal, dynamic model provisioning, and an OpenAPI-powered docs route.

Gateway API

Expose a clean, OpenAI-style endpoint that validates API keys, proxies requests to your local model runtime, and records usage.

User Dashboard

Create and revoke API keys, monitor request counts, inspect recent traffic, and understand monthly token consumption.

Admin Panel

Review users, active keys, quota posture, and recent model activity from a single operational view.

User access

Log in and manage keys.

Users can create keys, inspect recent usage, and track monthly quota from a clean dashboard.

User dashboard

Operations

Supervise the platform.

Admins get a snapshot of users, keys, traffic, and quota posture without touching raw database records.

Admin dashboard

Integration

Share a stable API surface.

Third-party apps can integrate through an OpenAPI-described route and standard bearer token auth.

Explore docs