GPU Pricing

This is the pricing cluster: what a GPU actually costs to rent, broken down provider by provider and updated as rates move. Each post pulls a named provider's published on-demand and spot rates (AWS P5, Azure ND-series, GCP A3/A4, CoreWeave, Nebius, Crusoe, and a couple dozen more) and lines them up against Spheron's live pricing so the comparison isn't hypothetical.

A second thread runs through here too: API pricing versus self-hosting. When a frontier model cuts its per-token rate, that changes the token volume where renting your own GPU cluster starts winning, and these posts do that math with the actual API prices, not estimates.

The pillar post, GPU Cloud Pricing Comparison 2026, is the widest lens: H100 and B200 rates across 15+ providers in one table, refreshed regularly. Start there if you're shopping broadly. Drop into an individual provider post (Azure H100, Lambda Cloud, Runpod, Vast.ai) if you already have a quote in hand and want to know whether it's actually competitive. Every number here traces back to a provider's own rate card or Spheron's live pricing API, and we say when a figure is spot instead of on-demand.