Infrastructure
The clouds, orchestration and plumbing that turn silicon into something you can rent by the hour.
This is where the money leaks. Nearly every recurring bill a technical team complains about — GPU hours, CI minutes, log ingestion, cluster management — is billed at this layer, and nearly every one of them has a free engine underneath the paid convenience.
14 decisions · 59 verified alternatives · 13 categories
Every decision at this layer
Ranked by how self-sufficient the leading alternative is. The score is a published rubric, not an opinion — see methodology.
Docker Desktop → Podman + Podman Desktop
Fully open-source, daemonless containers with a polished cross-platform GUI.
Free / open-source
Modal → vLLM
The standard open inference engine — the thing under most serving platforms.
Free and open source. You pay only for the GPUs you rent or own.
Red Hat OpenShift → k3s
Certified Kubernetes in a single small binary — no platform tax.
Free, open-source; optional SUSE Rancher Prime support subscription
Astronomer (Astro) → Apache Airflow
The same engine Astro runs, self-hosted and free.
Free and open source. You pay for the cluster and the people who run it.
GitHub Actions → Woodpecker CI
A small, container-native CI server you can run on one box.
Free and open source. You pay only for the machine it runs on.
Sentry → GlitchTip
A lightweight, fully open-source, Sentry-SDK-compatible error tracker.
Free to self-host; optional hosted plans from $0 (1K events/mo) to $250/mo (3M events/mo)
Locked-Down Consumer Routers → OpenWrt
Open-source Linux firmware for your router.
Free / open-source (flash onto supported hardware)
Datadog → Grafana + Prometheus
The open-source observability standard.
Free / self-host
Dropbox → Nextcloud
A full self-hosted cloud: files, sharing, and more.
Free / open-source (self-host); paid support/hosting optional
New Relic → SigNoz
An OpenTelemetry-native, single-app APM and observability platform.
Free / self-host; optional managed cloud
Splunk → OpenSearch
The Apache-2.0 search and log-analytics platform (the Elasticsearch fork).
Free / self-host; managed options available from AWS and others
AWS (EC2 compute) → Hetzner Cloud
Two to four times the hardware per dollar, 20 TB of traffic included.
From ~€4.50/mo (2 vCPU / 4 GB, 20 TB EU traffic); US locations slightly higher
GoDaddy → Porkbun
Flat, honest domain pricing with free privacy — the community favorite.
~$11/yr for .com — the renewal matches the intro price; free WHOIS privacy and SSL
AWS GPU Instances → RunPod
H100s from about $1.99/hr, running in under a minute.
Per-second billing, no commitment. H100 around $1.99/hr, with cheaper Community Cloud capacity and a pricier Secure Cloud tier. Serverless scales to zero, so idle inference endpoints cost nothing. Verified 2026-07-30.
Categories in this layer
Where this sits in the stack
L2 below · Chips & silicon — The physical accelerators — GPUs, TPUs, custom inference silicon — and the fabs that make them.
L4 above · Models & tooling — The models themselves, plus the gateways, vector stores, frameworks and observability that make them usable.