Four reference architectures, each priced across all 10 providers in the catalog by the same engine as the live estimator. Open one to see the per-provider totals, the cost topology, and every assumption behind the number — then load it into the calculator and change whatever you disagree with.
Ideal for early-stage startups and production web apps with up to 25k monthly active users.
A resilient, scalable 2-tier setup featuring 2 compute instances for high availability, 250 GB managed Postgres database with automated backups, 500 GB object storage for media, and load balancing.
Engineered for high concurrent checkouts, product catalogs, and intensive flash sales.
High-availability architecture with 6 medium instances, Multi-AZ relational database with read replication capacity, 2 TB hot storage with global CDN egress, and redundant load balancing.
Multi-service containerized architecture with managed Kubernetes clusters and service mesh.
Managed Kubernetes cluster (EKS/AKS/GKE) with 8 worker nodes (4 vCPU / 16 GB RAM each), dedicated multi-AZ database, 5 TB object storage, and heavy data transfer across private services.
Compute and memory-optimized instances designed for low-latency LLM serving and vector retrieval.
High-throughput architecture featuring high-capacity compute instances, scalable object storage for model weights and dataset embeddings, managed database for metadata, and high bandwidth egress.