Methodology
How Cloudpylon collects, normalizes, verifies, and refreshes cloud pricing data.
Collection
Prices are collected from official provider pricing sources and public catalogs by our ingestion pipeline, then stored as raw observations before any transformation. We never accept scraped third-party numbers.
Normalization
Raw observations are converted into one schema: on-demand hourly and monthly rates, spot and preemptible rates where offered, committed-use discounts, vCPU, memory in GiB, architecture, and region metadata. Currency conversions use daily reference rates and always display alongside the USD source figure.
Freshness
The catalog records provider pricing and computes rankings from its normalized database. Provider terms can change without notice, so confirm final terms on the provider's official page before purchasing.
Verification
Automated checks compare sampled stored values against fresh provider responses and flag discrepancies for review. Statistical outliers - sudden jumps beyond expected ranges - hold back publication until confirmed.
Limitations
Provider pricing evolves without notice; regional taxes, license bundles, startup credits, and negotiated discounts are out of scope. Always confirm final terms on the provider's official page before purchasing.
Last reviewed
2026-08-22