Weekly Pricing Pulse

About Pulse

The Weekly Pricing Pulse tracks what changed in published GPU cloud pricing across AWS, Azure, GCP, OCI, CoreWeave, Lambda, Vultr, Nebius, and Crusoe across the latest verified comparison period. AIForge Works compares each new harvest with the prior published verified harvest and highlights buyer-relevant movement: comparable on-demand price changes, New Silicon, restored / expanded visibility, newly absent exact provider + region + instance/SKU observations, and likely price/SKU rolls.

This is not a live quote, workload-specific recommendation, or personalized account analysis. For workload-specific recommendations, use Cloud Advisor. To inspect neighboring SKUs, regions, or commitment terms, open the Multi-Cloud Pricing Analyzer. To compare your own GPU spend against a comparable floor, use GPU Audit.

Editor's note

Pulse update - August 11, 2026: During our post-publication review, we found that the release build did not carry forward the Azure meter-classification evidence file used by the comparison. Azure pricing was present in both harvests, but 1,209 matched Azure observations were excluded from the initial movement-coverage count. We corrected that count from eight providers to all nine providers. No Azure source price, catalog floor, or OCPI value changed. The corrected comparison contains 13,807 current public on-demand GPU offers, 13,741 prior offers, and 13,496 exact provider + region + instance/SKU matches. This correction is a result of the collection and normalization enhancements described below and is part of expanding AIForge Works coverage and the beta hardening concluding this month.

This issue covers 14 days instead of the usual seven, as we elected not to publish last week. In the normalization process last Monday, we determined additional filter tuning was required. It took a bit longer than we had forecasted to update our cloud providers catalog collection and normalization processes. We addressed a large normalization filter gap in AWS bare-metal instances, which our collector had been excluding because AWS classifies them under a different product family than its virtual machines.

That correction is why this issue shows 690,186 more raw price observations than the last, a 12.4% increase, and 83,954 additional normalized pricing rows, up 10.2%. Nearly all of this is measurement correction: the same market, measured more completely. The restored T4 entries are not new silicon. We also separated 184 Lambda regional observations and three Vultr L40S observations caused by changes in our region-evidence classification. In addition, we repaired GCP coverage for 10 A3 High H100 regions and three existing Iowa accelerator-optimized shapes. Those 13 GCP catalog observations now appear in our pricing coverage but are excluded from New Silicon because they are measurement correction, not evidence of a GCP launch.

We have separated these corrections from genuine catalog movement throughout this issue, so that a fix on our side never reads as news on the providers' side. One distinct regional-offering signal remains: G5/A10G listings appeared in EU (Paris) and Europe (Spain) for the first time in our records. AWS's regional offerings API confirms all eight G5 sizes in both regions. This confirms regional offering, not currently available capacity or the date AWS introduced them. No AWS instance pricing shown in AIForge Works was inaccurate; the comparison set was missing SKUs. We would rather miss an issue than publish a number we cannot stand behind. Weekly cadence resumes with this August 10 issue, and the next Pulse is scheduled for Monday, August 17.

Ornn OCPI context

OCPI is the Ornn Cloud Price Index, a transaction-based benchmark for observed GPU cloud pricing that AIForge Works uses as market context alongside public provider catalog floors. This Pulse uses the latest reviewed Ornn x ICE OCPI comparison, through the August 9, 2026 20:00 UTC reading set. The August 10, 2026 OCPI artifact has not been promoted yet, so this section remains directional index context, not a current-week OCPI reading claim and not part of the weekly on-demand movement diff below.

Week at a glance

Among all 9 providers with comparable public on-demand GPU offers in both harvests, none had a price move over the 1% reporting threshold. AWS, CoreWeave, Lambda, Nebius, and Vultr changed catalog visibility through New Silicon, restored / expanded visibility, and newly absent exact catalog observations, shown below.

These movement counts apply only to comparable public-catalog on-demand prices. OCPI is a separate transaction-based market index and is shown independently.

• Across all 9 providers, no comparable public on-demand GPU price changes cleared the 1% reporting threshold.

Provider roll-up

No buyer-visible GPU catalog additions, removals, or public on-demand price moves above 1% were found this week for Azure, GCP, OCI, and Crusoe.

GPU movement coverage. All 9 providers had comparable public on-demand GPU offers in both harvests. Azure contributed 1,209 current offers and 1,209 prior offers; all 1,209 exact provider + region + instance/SKU observations matched across the comparison period.

Underlying provider feeds also shifted: Azure (+17,792 catalog records), OCI (-1,092 catalog records). After normalizing to comparable GPU offers, purchase terms, and verified hardware specs, those feed-level changes did not alter the public on-demand GPU offers reported in this Pulse.

Offers seen in earlier verified harvests returned to view this week: Lambda (21) and Vultr (1). Because these provider and SKU combinations appeared in earlier verified harvests, we treat them as restored catalog visibility rather than brand-new silicon launches.

Catalog changes this week

Catalog change counts summarize week-over-week visibility changes in the compared harvests. They show provider + region + instance/SKU observations that became newly visible or no longer visible between harvests. We do not describe these as confirmed capacity launches or retirements unless provider-published sources say so.

Measurement corrections

13 newly represented GPU catalog observations were held out of New Silicon across GCP (13). These rows reflect a verified collector coverage repair, not evidence that the provider launched new hardware or capacity during this comparison period.

New Silicon

96 New Silicon GPU catalog entries were detected across AWS (96). 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 no-longer-visible observation. It still does not by itself prove net-new capacity or a brand-new accelerator family.

Restored / expanded visibility

22 previously seen public GPU catalog observations were visible again this week across Lambda (21), Vultr (1).

Each provider and SKU had appeared in an earlier verified comparison; the region shown is where the offer was visible this week. We classify these as restored or expanded catalog visibility, not a new hardware launch, proof of added capacity, or confirmation that the offer is currently orderable.

Newly absent exact catalog observations

61 newly absent exact provider + region + instance/SKU GPU catalog entries were detected across CoreWeave (13), Lambda (25), Nebius (4), Vultr (19). Entries are grouped by provider so high-volume catalog churn does not hide smaller-provider changes.

A newly absent exact catalog observation means the provider + region + instance/SKU combination was present in the prior harvest but absent from this one in this week's public-catalog comparison, even when a same-instance replacement looks likely. Multiple regional observations can refer to the same SKU string; absence here does not prove the SKU is unavailable to buy.

Top on-demand movers by provider

No comparable offers met the 1% movement threshold for this section.

Methodology

Evidence through. Pricing evidence in this issue runs through August 10, 2026. This evidence-through date is recorded separately from the publication date shown in the article header.

What we compared. This Pulse compares the last published verified harvest (July 27, 2026) against the current harvest (August 10, 2026). The comparison spans 14 days; it is not characterized as a single seven-day week.

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. Providers with zero comparable GPU offers in either week are excluded from GPU price-movement conclusions. 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 no-longer-visible observation and has not been seen in the lookback window. Restored / expanded visibility means the provider and instance/SKU appeared in an earlier verified harvest and is visible again, sometimes in additional regions. A newly absent exact catalog observation means the exact provider + region + instance/SKU combination was present in the prior harvest but absent from this one; multiple regional observations can share one SKU string.

Scope and limits. The generated Pulse diff certifies on-demand movement only. All prices are in US dollars. Catalog entries missing a clean price or verified GPU specs are dropped before comparison. This Pulse compiles publicly published prices and is not a live quote, availability guarantee, or workload-specific advice.

This week's coverage. The August 10, 2026 harvest completed across 9 providers. Across the broader catalog, 6,275,283 raw price observations became 907,528 normalized pricing rows, up 83,954 (+10.2%) week over week. Within that normalized catalog, instance rows increased by 2,663 (+2.4%) to 112,558. The supporting instance-spec table grew to 2,485,539 rows (up 223,495 (+9.9%)). This Pulse narrowed the current catalog to 13,807 normalized comparable public on-demand GPU offers, up 66 (+0.5%). Of those, 13,496 exact provider + region + instance/SKU observations were present in both harvests and eligible for direct movement comparison.

Next Pulse: Week of 2026-08-17.