Cloud-9 finds the cheapest, most stable spot instances for your workload — recommends the right Savings Plan commitment, generates Karpenter configs, and deploys to your ASG in one click. No agents, no 2-week setup.
Drag the slider — most teams land somewhere in between.
Real-time scan across all regions and AZs to find the absolute price floor for your specs.
Weighted index of price, stability, and interruption risk across hardware generations.
Combines AWS Spot Advisor interruption-rate data with a next-day price-trend model to flag pools drifting toward tighter capacity — the real spot risk, not a stock-ticker price guess.
Shaved Ice algorithm sizes your EC2, Fargate, and Lambda Savings Plan commitments to minimise total cost across all compute types.
Generates production-ready Terraform for mixed spot/on-demand ASGs — with workload eligibility gating, FIS interruption drills, and a revert-safe on-demand floor. Ships as a ready-to-open pull request: savings estimate, eligibility evidence, and revert path in the PR description, so the change flows through your CI/CD with a human in the loop.
Generates a scored, interruption-weighted NodePool + EC2NodeClass YAML for Karpenter v1 — pre-tuned for spot-to-spot consolidation with ≥15 instance types.
Same workload, priced across AWS spot, AWS on-demand, Azure spot, and Hetzner dedicated — one table, no accounts needed on the other clouds.
Peer-reviewed at ICPE '25, our algorithm treats cloud capacity as a high-frequency liquid market — then executes on the insight, not just reports it.
567K+ price points ingested from the AWS Spot feed across 18 regions and all availability zones.
Historical interruption rates are mapped against real-time fleet demand signals to produce a composite score.
Your choice of control: Copilot generates Terraform for review, the PR export pushes it through your own pipeline — repo stays the source of truth — or Autopilot deploys directly via a scoped cross-account role with a full audit trail.
“Shaved Ice: Optimal Compute Resource Commitments for Dynamic Multi-Cloud Workloads”
Cloud-9's Commitment Advisor implements the Shaved Ice algorithm from this peer-reviewed paper — minimising total cost across spot, on-demand, and Savings Plan commitments.
Read the full paper →M. Stokely · N. Nadgir · J. Peele · O. Kostakis
Cloud-9 was born out of research at the Technical University of Munich. We build the tools we wished existed when we were managing AWS bills ourselves.
No credit card required. Free tier includes spot finder, value scorer, and price trends.
Get Started — Free