InferiaLLM

Quickstart

Get up and running with Inferia LLM

You can run Inferia LLM using our Python package (recommended) or via Docker.

1. Using Python Package

The easiest way to get started is by installing the inferiallm package.

Prerequisites

  • Python 3.10+
  • PostgreSQL 14+
  • Redis 6+

Installation

  1. Install the package:

    python3 -m venv .venv
    source .venv/bin/activate
    pip install inferiallm
  2. Configure Environment:

    curl -o .env https://raw.githubusercontent.com/InferiaAI/InferiaLLM/main/.env.sample

    Edit the .env file to set your database credentials and secrets.

    Critical Variables:

    VariableDescription
    DATABASE_URLPostgreSQL connection string (e.g., postgresql://inferia:inferia@localhost:5432/inferia)
    REDIS_HOSTRedis host (default: localhost)
    REDIS_PORTRedis port (default: 6379)
    JWT_SECRET_KEYSecret for signing access tokens (generate a long random string)
    INTERNAL_API_KEYSecret for service-to-service authentication
    SECRET_ENCRYPTION_KEY32-byte base64 key for encrypting provider credentials
    SUPERADMIN_EMAILEmail for the initial admin account
    SUPERADMIN_PASSWORDPassword for the initial admin account
  3. Initialize the Database: This creates the necessary tables, roles, and initial admin user.

    inferiallm init
  4. Start Services: Run all gateways and the dashboard in a single command.

    inferiallm start

    Access Points:

    ServiceURL
    Dashboardhttp://localhost:3001
    Orchestration APIhttp://localhost:8080
    Filtration Gatewayhttp://localhost:8000
    Inference Gatewayhttp://localhost:8001

2. Using Docker Compose

If you prefer containerization, use the deploy directory.

Prerequisites

  • Docker & Docker Compose

Steps

  1. Pull Docker Image:

    docker pull inferiaai/inferiallm:latest
  2. Configure Environment: Edit .env with your secrets. The Docker Compose file will use these variables.

    curl -L https://raw.githubusercontent.com/InferiaAI/InferiaLLM/main/.env.sample -o .env
    nano .env

    [!IMPORTANT] At minimum, set JWT_SECRET_KEY, INTERNAL_API_KEY, and SECRET_ENCRYPTION_KEY.

  3. Start Services:

    docker run -d \
      --name inferia-app \
      --env-file .env \
      -p 8000:8000 -p 8001:8001 -p 8080:8080 -p 3000:3000 -p 3001:3001 \
      inferiaai/inferiallm:latest
  4. Access Dashboard: Open http://localhost:3001.

    Default credentials: admin@example.com / admin123

Docker Services

The unified Docker image starts these services:

ServicePortDescription
Filtration Gateway8000Auth, RBAC, Guardrails
Inference Gateway8001OpenAI-compatible API
Orchestration Gateway8080Compute & Deployment Management
DePIN Sidecar3000Nosana/Blockchain interactions
Dashboard3001Admin UI

On this page