Skip to main content

Deployment

Deploy LlamaFarm locally, in containers, or on clusters.

Philosophy​

  • Write once, deploy anywhere
  • Local first, cloud ready, edge capable, production grade

Quick starts​

# Local
llamafarm up
llamafarm up -f llamafarm.dev.yaml
llamafarm up --watch

# Docker
llamafarm build --docker
docker run -p 8080:8080 llamafarm:latest

docker run --gpus all -p 8080:8080 llamafarm:latest

# Cloud (shortcuts)
llamafarm deploy aws --region us-east-1
llamafarm deploy azure --resource-group my-rg
llamafarm deploy gcp --project my-project

Continue with:

  • Docker Compose → /deployment/docker-compose
  • Kubernetes → /deployment/kubernetes

See also production best practices (HA, monitoring, security, scaling) in the sections below.