Weekly Pricing Pulse

About Pulse

The Weekly Pricing Pulse presents what changed in published GPU cloud pricing across the 9 tracked providers in the last 7 days. Every week, AIForge Works runs a fresh harvest across AWS, Azure, GCP, OCI, CoreWeave, Lambda, Vultr, Nebius, and Crusoe, diffs the new data against the prior week, and surfaces the moves worth knowing about: biggest price changes on existing comparable entries, New Silicon, Deprecated SKUs, and likely price/SKU rolls.

What the Pulse is NOT: it is not a live quote, and it is not workload-specific advice. The numbers reflect public catalog prices at the moment of harvest. For a live quote, check the provider. Within AIForge Works, use Cloud Advisor for workload-specific recommendations, open the Multi-Cloud Pricing Analyzer to see neighboring SKUs, regions, or commitment terms for any mover that catches your eye, and use GPU Audit to compare your own GPU spend against a comparable floor.

Editor's note

June 1 is an Azure-pricing week, not a broad-market repricing week. All comparable on-demand movement above the 1% threshold came from Azure, with the clearest rows showing V100 and A100 public-rate changes while the rest of the provider set mostly shifted in catalog visibility. AWS, GCP, Lambda, and Vultr showed newly visible or no-longer-visible GPU catalog entries; OCI, CoreWeave, Crusoe, and Nebius were quiet in this weekly comparison. We are keeping the headline table conservative: one Azure H100 price looked like a secondary billing meter rather than a full-compute instance price, and a few other rows did not have enough market coverage to treat as strong signals. For buyers, the takeaway is to refresh Azure assumptions before carrying forward prior-week A100 or V100 comparisons, and treat catalog-change sections as a watchlist rather than proof of new physical capacity.

What moved this week

All qualifying on-demand price movement this week was in Azure. AWS, GCP, Lambda, Vultr also changed catalog visibility through New Silicon and Deprecated SKU entries, shown below. Same-instance SKU rolls are tagged as likely price changes.

• Azure V100 32 GB increased 100.0% in swedencentral, on a per-GPU basis.
• Azure A100 80 GB increased 8.2% in australiaeast, on a per-GPU basis.
• Azure A100 80 GB decreased 7.6% across several regions, on a per-GPU basis.

Top on-demand movers by provider

Azure

High-confidence Azure movers, ranked by confidence and then percentage change. (33 Azure price entries moved >1% in total; one ND96is H100 partial-meter row and lower-confidence rows are held out of the headline table.)

Lower-confidence / under-review observations: one Azure Standard_ND96is_H100_v5 southcentralus row is excluded because it matches the partial-meter pattern later found in the ND96is audit. Thin K80/M60/A10 rows and uncovered Radeon V710 rows are omitted from the headline table even though they remain in the raw diff artifact.

Up to 5 price observations per provider, ranked by confidence first, then percentage change. See Methodology for how per-GPU rates are computed.

Provider roll-up

No change this week: OCI, CoreWeave, Crusoe, Nebius.

Existing price changes track comparable provider + region + instance/SKU observations over the 1% movement threshold. New Silicon excludes same-provider, same-geo, same-instance likely SKU rolls. Deprecated SKUs remain exact catalog SKUs no longer visible as purchasable rows.

Catalog changes this week

Catalog change counts are artifact-level visibility changes. They show provider + region + instance/SKU observations that became newly visible or no longer visible between harvests. Provider-level add/remove churn exists for Lambda (27 newly visible, 9 no longer visible); Vultr (5 newly visible, 3 no longer visible), but this comparison did not find same-provider, same-geo, same-instance pairs. We do not describe these as likely SKU rolls unless that exact overlap is present.

New Silicon this week

54 New Silicon on-demand GPU catalog entries were detected across AWS (17), GCP (5), Lambda (27), Vultr (5). Entries are grouped by provider so high-volume catalog churn does not hide smaller-provider changes.

New Silicon means a provider + region + instance/SKU observation newly visible in this harvest that is not paired with a same-provider, same-geo, same-instance deprecated SKU. It still does not by itself prove net-new capacity or a brand-new accelerator family.

Detailed New Silicon rows are held back for source-backed review this week because one AWS catalog family needs GPU-label disposition before we characterize it publicly. Provider-level counts remain in the roll-up above; we are not treating those entries as proof of new physical capacity or a new accelerator family.

Deprecated SKUs this week

13 Deprecated SKU on-demand GPU catalog entries were detected across Azure (1), Lambda (9), Vultr (3). Entries are grouped by provider so high-volume catalog churn does not hide smaller-provider changes.

Deprecated SKU means the exact provider catalog SKU was present in the prior harvest but absent from this one. It is no longer purchasable as that exact SKU in the artifact comparison, even when a same-instance replacement looks likely.

Methodology

What we compared. This Pulse compares last week's harvest (May 25, 2026) against this week's (June 1, 2026). Both harvests run automatically against the same 9 cloud providers.

How we compare prices fairly. Different providers package GPUs differently: some sell single-GPU instances, others sell 8-GPU instances. To compare apples-to-apples, we divide every instance price by the number of GPUs on that instance. All movement percentages in this Pulse are based on that per-GPU hourly rate, not the raw instance price.

What counts as a real move. We only flag price changes greater than 1%. Anything smaller is treated as rounding noise and excluded.

Reserved/committed terms. The generated Pulse diff certifies on-demand movement only. If the editor's note discusses 1-year or 3-year reserved or committed terms, that claim must be reviewed against separate evidence before publication.

Currency. All prices are in US dollars. Catalog entries missing a clean price or verified GPU specs are dropped before comparison.

How we choose summary callouts. The short summary only highlights price moves that are broad enough to read as a market signal. Narrow one-provider movements may still appear in detail tables, but we do not treat them as market-wide repricing.

This week's coverage. All 9 providers reported clean harvests. After normalizing 5.51M raw catalog price observations against verified hardware specs, the broader MPA catalog includes all deployable instance offers. This Pulse narrows that catalog to roughly 14,000 comparable on-demand GPU observations for week-over-week movement analysis.

This Pulse compiles publicly published prices. It is not a live quote and is not workload-specific advice. See the About section at the top for links to AIForge Works pricing and availability tools.

Next Pulse: Week of 2026-06-08.