Plan your infrastructure
Pick a stage, or pick layers by hand. Each one gets a suggested starting provider — the cheapest that clears free-tier-or-a-clean-price — with the real reason, not a generated one, and a link to the full category if you want a different pick.
Keyword match against words like "postgres", "file uploads", "background jobs" — not a model call. Hosting is always included; every project runs somewhere.
Picks inetGeek's usual starting checklist for that stage — every layer below stays yours to add or remove.
Only changes the Auth and Secrets management picks — the only layers where any provider documents these today. A requirement no candidate satisfies says so, rather than picking one anyway.
$0/mo
This is not what you will pay. It's the sum of each picked layer's cheapest documented starting point — a real free tier counted as $0, or a minimum paid plan's own quoted price. Traffic, storage and usage move every number here upward the moment a real project runs, and layers with no clean published starting price (usage-based pricing with no flat minimum) are left out of this total below, not counted as free.
Front-end apps deployed from Git onto AWS with usage-based pricing and no monthly minimum, including 1,000 build minutes and 15 GB of transfer under the free tier.
Teams who want more than one open-source data service from one vendor and one bill — Postgres beside Kafka, ClickHouse or OpenSearch — with the availability tier chosen per service rather than fixed.
Workloads already inside AWS, where keeping data in the same region as the compute avoids the transfer charges that dominate an object storage bill.
Projects that fit inside the free 25,000 monthly active users, which includes an enterprise connection, self-service SSO and SCIM at no cost.
EU-resident sending where you want to choose how long message data is kept — retention is a setting with a floor of zero on paid plans, not a policy you inherit.
Small projects that want a real telemetry platform for nothing — 500 GB a month loaded with 30-day retention, permanently, is the most generous free allowance in this category.
AWS-resident systems that want a queue with a permanent free million requests a month and no capacity to plan — SQS scales without a throughput setting to choose.
Teams already on AWS that want CDN bundled with other AWS services under one flat monthly price rather than itemized per-GB billing.
Teams already on AWS that want engine choice (Valkey, Memcached or Redis OSS) under one billing model, or that specifically need Valkey's newer durable-persistence option.
Teams that want a fully managed vector search product with hybrid search and metadata filtering built in, and don't want to operate the underlying index themselves.
Teams whose secrets access is dominated by CI/CD and service identities rather than human users — those ride free, so cost tracks headcount, not automation volume.
Teams that already run their own compute and want a CI product built around that from the start, rather than one that treats self-hosting as an escape hatch from a hosted-minutes model.
Teams that want per-token APIs and dedicated GPUs from one vendor with no plan fee — minute-granularity billing means a model that runs for ten minutes costs ten minutes.
Training runs that want a whole HGX node and move large datasets — free egress and free I/O operations remove the charge that usually dominates a data-heavy job.
Pick a stage or a layer above to start.