Consilium
Back to Documentation

Platforms – cloud and Docker

Related

Prerequisites

  • Docker and Docker Compose v2+
  • 2 GB RAM minimum (4 GB recommended for production)
  • API keys for at least one LLM provider (OpenAI, Anthropic, Google, Groq, or xAI)

    • Quick Start (Docker)

    git clone https://github.com/skadri1601/Consilium.git
cd Consilium
cp .env.example .env
# Edit .env with your API keys and POSTGRES_PASSWORD
docker compose -f docker-compose.selfhost.yml up

    Once running, open http://localhost:3000 in your browser.

    Services

    Startup Order

    PostgreSQL and Redis start first with health checks. Agents waits for Redis. API waits for PostgreSQL, Redis, and Agents. Web waits for API.

    Environment Variables

    Core

    Authentication

    LLM Provider Keys

    You need at least one provider key. The engine automatically routes to available models.

    Frontend

    Observability

    Payments

    Manual Setup (Without Docker)

    # Install dependencies
pnpm install

# Generate Prisma client
pnpm --filter @consilium/database exec prisma generate

# Push schema to database
pnpm --filter @consilium/database exec prisma db push

# Start all services
./run.sh

    run.sh starts: Web on port 3000, API on port 4000, Agents on port 8000.

    Health Checks

    All Docker services have built-in health checks with interval: 10-15s, retries: 5, and start_period: 10-30s.

    Production Recommendations

    Port Overrides

    All ports are configurable via environment variables: