Compare
SIAX vs Google Cloud: price and EU residency
Google Cloud is the hyperscale cloud that's strongest where data and machine learning meet. BigQuery is the single clearest reason to choose GCP: a data warehouse where storage and compute are separated, that scales to petabytes without anyone having to size a cluster, and that answers SQL over enormous datasets in seconds. We have no equivalent. Managed Postgres at 89 kr/month goes a lot further than people expect, but it's a different product category.
The same goes for ML. Vertex AI, the Gemini models, the TPUs, and the ready-made pipelines for training and deployment are a coherent toolbelt we don't try to compete with. Our AI gateway at 149 kr/month is a routing layer in front of model providers with cost caps, tracing, and EU residency for the traffic — useful, but not the same thing as a training platform. GKE, including Autopilot, is also among the best managed Kubernetes available, and Google's global network with anycast load balancing is effectively a competitive advantage in its own right.
What GCP shares with the other hyperscalers is the price level for small workloads, a usage-based invoice that requires active management, and the lock-in that results when the core of the architecture is built on services that only exist there.
Where SIAX is stronger
- The price level for ordinary workloads. An app plus managed PostgreSQL costs 188 kr/month with us. The equivalent Cloud Run plus Cloud SQL setup typically lands several times higher, and prices change.
- Predictability. Fixed price per service instead of usage that has to be tracked with budget alerts and cost reports.
- EU data residency without exception, with a Swedish company. All data, all logs, all backups in the EU.
- No lock-in. Everything we run is open source: PostgreSQL, S3-compatible storage, containers, Git, and CI. If we shut down, you can keep running the same software elsewhere — that's a promise no hyperscaler can make about its own proprietary services.
- Support in Swedish, from the people who actually run the platform, with no support contract priced separately.
Where Google Cloud is stronger
- BigQuery. A serverless, petabyte-scale data warehouse with no operations overhead, with separated storage and compute. It has no equivalent with us or with most others.
- Vertex AI, the Gemini models, and the TPUs. A coherent toolset for training, fine-tuning, and deploying models — not something we offer.
- GKE and Autopilot. Top-tier managed Kubernetes, with a control plane and an upgrade model that few come close to.
- The global network. Anycast load balancing, low latency worldwide, and regions on every continent.
Side by side
| SIAX | Google Cloud | |
|---|---|---|
| Price for a small production app | 99 kr/month for the app, 89 kr/month for Managed Postgres. Fixed price. | Cloud Run plus Cloud SQL with a small instance typically lands several times higher per month. Depends on traffic and configuration, and prices change. |
| Analytics and data warehousing | PostgreSQL goes a long way for operational analytics, but we have no equivalent to BigQuery. | BigQuery: serverless, petabyte scale, separated storage and compute. Their strongest card. |
| AI and machine learning | AI gateway from 149 kr/month that routes to model providers with cost caps, tracing, and EU residency. We don't train models and have no GPU or TPU fleet to sell. | Vertex AI, Gemini, TPUs, ready-made training and deployment pipelines. |
| Kubernetes | Container hosting on open source. No managed Kubernetes control plane in GKE's class. | GKE and Autopilot — among the best managed Kubernetes available. |
| Outbound traffic | Included in the plan. | A limited amount of free outbound traffic per month, then per GB, with price differences between destinations. |
| Geographic reach | EU data centers in Germany and Finland. No presence outside Europe. | Regions on every continent with a global anycast network. |
| Certifications | None. Neither ISO 27001 nor SOC 2. GDPR compliance, a data processing agreement, EU residency, and open source as auditability. | A full battery of certifications including ISO 27001 and SOC 2, with evidence packages for security reviews. |
| Lock-in | Open source all the way. Built so you can leave us without rewriting anything. | BigQuery, Spanner, Firestore, and Pub/Sub have no self-hosted equivalents. A move becomes a rewrite. |
| Invoice and support | Fixed monthly price, support in Swedish included. | Usage-based invoice that requires monitoring. Technical support with an SLA is priced separately. |
Choose Google Cloud if…
Choose Google Cloud if data or machine learning is the core of the product. If you need to run analytical queries over billions of rows, build a data platform with Dataflow and Pub/Sub, or train and deploy your own models — then there's no honest comparison to make, because we don't have those services. Choose GCP too if you run Kubernetes at scale and want someone else managing the control plane, if you need global load balancing with low latency across multiple continents, or if your procurement process requires the vendor to hold certifications like ISO 27001 and SOC 2. We have no certifications, and that's a legitimate reason to rule us out. Also expect that a move away from BigQuery, Spanner, or Firestore isn't a move but a rewrite — that's an argument for choosing correctly from the start, not for choosing us.
Choose SIAX if…
Choose SIAX when the workload is ordinary: web apps, APIs, PostgreSQL, files, jobs, and queues, with users in the Nordics or Europe. That's where the price difference is greatest and where GCP's particular strengths don't do you any good. You get an invoice that looks the same every month, EU data residency with no caveats, and a stack you can lift out. A common and sensible setup is to run the base infrastructure with us and keep BigQuery or Vertex AI with Google for what they're good at — we have no interest in pretending you have to give them up.
See products and pricing