AutoSpotting vs commercial Spot managers

What a commercial Spot manager does well

Commercial Spot managers such as Spot.io are mature products. They automate the hard parts of running Spot: selecting and diversifying instance types, replacing capacity when it is reclaimed, falling back to on-demand, and presenting a dashboard over your fleet. If your priority is a managed service with a vendor operating the automation for you, that model is a reasonable fit, and we will not pretend otherwise on this page.

AutoSpotting automates the same core job, converting AutoScaling group capacity to Spot with on-demand fallback and instance-type diversification. Where the two differ is the deployment model, so the honest comparison is on architecture, ownership, and pricing model rather than on a feature checklist. For current terms and capabilities of any specific vendor, check their own site directly; we do not restate other companies' features or pricing here.

The architectural difference: your account vs a SaaS backend

A commercial SaaS Spot manager runs a backend outside your account and connects to your AWS environment with cross-account access to manage capacity on your behalf. That is how the managed model works, and for some teams the convenience is worth it.

AutoSpotting runs entirely inside your own AWS account, as Lambda functions, with no SaaS backend and no cross-account access to grant. Nothing about your infrastructure is transmitted to an external service, and it uses minimal IAM permissions. This is the design AutoSpotting has kept since it was first built in 2015-2016, when its author evaluated an early SaaS Spot manager and preferred not to convert everything to a vendor-specific construct or grant a third party standing access to the account.

Deployment model and lock-in

Commercial managers generally onboard your workloads into their own group constructs. That gives them a consistent surface to manage, and it also means the shape of your setup follows the vendor, which can make leaving harder later. AutoSpotting works from a tag on the AutoScaling groups you already run. There is nothing to migrate into, and removing the tag reverts the group to pure on-demand.

Open-source core and pricing model

The AutoSpotting core has been open source since 2016 and is available as the Community Edition under the OSL-3.0 license, so you can read the code, build it yourself, and run it at no cost. The commercial edition adds tested binaries, support, and the latest enhancements, priced at 10% of the savings it generates on your AWS bill, billed through the AWS Marketplace, with a free tier for small installations. It is usage-based: you pay a share of realized savings rather than a seat or platform fee, and there is no long-term commitment.

A note on forks of the open-source core

Because AutoSpotting's core has been open source since 2016, some commercial products trace back to that early codebase. Xosphere, for example, is a commercial fork of the original MIT-licensed AutoSpotting. AutoSpotting is the actively developed original: its architecture moved from the initial cron-based model to an event-based one years ago. If you are evaluating a fork, it is worth confirming which architecture it runs today and comparing that against the current AutoSpotting directly, rather than assuming a shared history means shared behavior.

Which model fits you

If you want a fully managed SaaS and are comfortable with a vendor operating the automation and holding cross-account access, a commercial Spot manager is a reasonable choice, and you should compare its current features and pricing on its own terms. If you want the automation to run inside your own account with no SaaS backend, an open-source core you can inspect, tag-based setup on the groups you already have, and usage-based pricing tied to realized savings, that is what AutoSpotting is built for.