The public on-demand rate of AI compute, captured across the locked twenty-seven provider headline panel. The anchor configuration is NVIDIA H100 SXM 80GB. United States dollars, per GPU-hour.
| Index | Value | Basis |
|---|---|---|
| SVCI Public GPU Rate Index | 103.5 | Headline composite, base = Issue 001; weights 0.45 / 0.45 / 0.10 |
| Compute Opacity Index | 28.5 | 100 times (1 minus Composite Coverage) |
| Hyperscaler Premium | 2.7352x | Tier 1 median over the pooled non-hyperscaler median |
| Neocloud Discount | 0.6344 | Tier 2 median relative to the Tier 1 median |
| Price Floor | 2.97 | Median of the lowest ceil(N times 0.25) included headline rates |
| Coverage Score | 0.7150 | Weighted panel coverage |
The Confidence Score remains unpublished pending the per-row confidence audit. All other readings are calculated from the Issue 006 data layer.
Receive the benchmark instead of hunting for it. The public issue stays open; the subscription delivers the Compute Price Brief, the market reading that explains it, and buyer tools as they launch.
| Tier | Weight | Panel slots | N in window | Median USD |
|---|---|---|---|---|
| Tier 1, Hyperscalers | 0.45 | 4 | 4 | 10.53065 |
| Tier 2, GPU clouds and neoclouds | 0.45 | 18 | 9 | 3.85000 |
| Tier 3, Marketplaces and aggregators | 0.10 | 5 | 2 | 4.47000 |
Tier 1 and Tier 2 medians are unchanged from Issue 004. Tier 3 decreases from $4.62 to $4.47.
Hyperstack increased its exact public on-demand H100 SXM rate from $2.40 to $3.20. The rate remains below the Tier 2 median, so the Tier 2 median stays at $3.85.
RunPod lowered the controlling Secure Cloud H100 SXM rate from $3.29 to $2.99. That change lowers the two-observation Tier 3 median from $4.62 to $4.47.
Up $0.80, or 33.3%. No Tier 2 median effect.
Down $0.30, or 9.1%. Tier 3 median decreases to $4.47.
Issue 006 resolves fifteen of twenty-seven providers to comparable fixed public H100 rates. Twelve are captured and excluded because the public surface is a floor, dynamic marketplace, aggregator pass-through, sales gate, account-gated listing or incompatible product mode.
| Status | Count |
|---|---|
| Headline-eligible public rates | 15 |
| Captured and excluded | 12 |
| Pending | 0 |
| Total headline panel | 27 |
| Tier | Resolved | Coverage |
|---|---|---|
| Tier 1, Hyperscalers | 4 / 4 | 100.0% |
| Tier 2, GPU clouds | 9 / 18 | 50.0% |
| Tier 3, Marketplaces | 2 / 5 | 40.0% |
| Weighted Coverage Score | 0.7150 | |
| Compute Opacity Index | 28.5 |
Issue 006 follows an interrupted cycle. No provider row is carried forward solely from a prior issue. Fresh captures or disclosure-status observations establish a new evidence baseline for all twenty-seven panel entries.
Issue 005 is preserved separately as a sealed release-candidate archive record. The Issue 006 public movement comparison remains stated against Issue 004, the previous published issue, as directed by the working capture ledger.
Atlas Cloud: the exact $2.95 public list price is accepted. Account-state UI is excluded from public evidence.
RunPod: Secure Cloud on-demand H100 SXM controls under F009. Community Cloud at $2.69 is supplementary.
Paperspace: the official $5.95 on-demand rate controls. The $2.24 three-year commitment rate is excluded, and promotional watch remains active.
Tier 3 remains a two-of-five fixed-rate observation. RunPod and Paperspace enter the tier median. Vast.ai, TensorDock and Shadeform do not because dynamic marketplace rates, host-varying floors and aggregator pass-through prices are not comparable fixed provider rates. The 0.10 weight remains locked.
The canonical data layer, calculation report, capture ledger, review log and selected public evidence manifest are published with the issue. Raw captures and account-state exploration are not included in the web deployment.
The headline panel contains twenty-seven provider slots. Fifteen comparable fixed public H100 rates enter the medians. Twelve captured rows are disclosed as non-inclusions. No row remains pending.
| ID | Provider | Status | Median | Rate | Reason |
|---|---|---|---|---|---|
| E001 | AWS | Priced | Yes | $6.88 | on_demand_fixed |
| E002 | Azure | Priced | Yes | $12.29 | on_demand_fixed |
| E003 | GCP | Priced | Yes | $11.0613 | on_demand_fixed |
| E004 | Oracle | Priced | Yes | $10.00 | on_demand_fixed |
| ID | Provider | Status | Median | Rate | Reason |
|---|---|---|---|---|---|
| E005 | CoreWeave | Priced | Yes | $6.155 | on_demand_fixed |
| E006 | Crusoe | Priced | Yes | $3.90 | on_demand_fixed |
| E007 | Lambda | Priced | Yes | $3.99 | on_demand_fixed |
| E008 | Nebius | Priced | Yes | $3.85 | on_demand_fixed |
| E009 | Together AI | Priced | Yes | $3.99 | on_demand_fixed |
| E010 | Hyperstack | Priced | Yes | $3.20 | on_demand_fixed |
| E011 | Verda | Priced | Yes | $3.25 | on_demand_fixed |
| E012 | Voltage Park | Captured, excluded | No | starting_floor_and_contact_pricing | |
| E013 | Fluidstack | Captured, excluded | No | no_public_pricing_surface | |
| E014 | Nscale | Captured, excluded | No | no_public_pricing_surface | |
| E015 | IREN | Captured, excluded | No | quote_gated_contact_sales | |
| E016 | ChronoScale | Captured, excluded | No | no_public_pricing_surface | |
| E017 | GMI Cloud | Captured, excluded | No | from_floor_and_official_surface_conflict | |
| E018 | CUDO Compute | Captured, excluded | No | quote_on_request | |
| E019 | Spheron | Captured, excluded | No | from_floor_marketplace_aggregator | |
| E020 | Jarvislabs | Priced | Yes | $2.69 | on_demand_fixed |
| E021 | Hyperbolic | Captured, excluded | No | no_public_comparable_fixed_rate | |
| E022 | Atlas Cloud | Priced | Yes | $2.95 | on_demand_fixed |
| ID | Provider | Status | Median | Rate | Reason |
|---|---|---|---|---|---|
| E023 | RunPod | Priced | Yes | $2.99 | secure_cloud_on_demand |
| E024 | Vast.ai | Captured, excluded | No | marketplace_dynamic | |
| E025 | Paperspace | Priced | Yes | $5.95 | on_demand_fixed |
| E026 | TensorDock | Captured, excluded | No | host_marketplace_floor | |
| E027 | Shadeform | Captured, excluded | No | aggregator_passthrough |