Simulate a cloud architecture (1-8 resources on AWS, GCP, Azure, OCI or DigitalOcean) under a traffic load and get modeled cost per hour, p50/p95/p99 latency where supported, throughput, error rate, reliability and autoscaling results in one synchronous call. Use it before provisioning or resizing to compare designs. No account or saved simulation; deterministic by seed. $0.003 per call.