# Sorso View Compute Index, Issue 004, Capture Ledger v1.2

Issue: SVCI Issue 004. Publication date: 2026-07-01. Package revision: v1.2, 2026-07-02.

v1.2 correction scope: public evidence crop and redaction only. No rates, inclusions, medians, SVCI, Coverage Score, or COI changed.

State: 27 captured, 15 priced, 12 captured-excluded, 0 pending. SVCI 103.8030, display 103.8. Coverage 0.7150. COI 28.5.

| ID | Provider | Tier | Status | Inclusion | Rate USD per GPU-hour | Billing mode | Source | Basis or flag |
| --- | --- | --- | --- | --- | ---: | --- | --- | --- |
| E001 | AWS | T1 | priced | Y | 6.88 | on_demand_fixed | https://aws.amazon.com/ec2/pricing/on-demand/ | none |
| E002 | Azure | T1 | priced | Y | 12.29 | on_demand_fixed | https://azure.microsoft.com/en-us/pricing/details/virtual-machines/linux/#pricing | none |
| E003 | GCP | T1 | priced | Y | 11.0613 | on_demand_fixed | https://cloud.google.com/products/compute/pricing/accelerator-optimized | none |
| E004 | Oracle | T1 | priced | Y | 10 | on_demand_fixed | https://www.oracle.com/cloud/price-list/ | none |
| E005 | CoreWeave | T2 | priced | Y | 6.155 | on_demand_fixed | https://www.coreweave.com/pricing | none |
| E006 | Crusoe | T2 | priced | Y | 3.9 | on_demand_fixed | https://crusoe.ai/cloud/pricing | none |
| E007 | Lambda | T2 | priced | Y | 3.99 | on_demand_fixed | https://lambda.ai/pricing | none |
| E008 | Nebius | T2 | priced | Y | 3.85 | on_demand_fixed | https://nebius.com/prices | none |
| E009 | Together AI | T2 | priced | Y | 3.99 | on_demand_fixed | https://www.together.ai/pricing | Issue 003 used the Together AI GPU Clusters product card at 4.79/hr per GPU. Issue 004 uses the top-level Together AI pricing page, GPU Clusters section, On-demand table, NVIDIA HGX H100 row at 3.99/hr. The gpu-clusters product-page H100 card still displays 4.79/hr per GPU and is retained as a public surface discrepancy. Dedicated Inference and reserved-duration prices are separate product modes and are excluded. |
| E010 | Hyperstack | T2 | priced | Y | 2.4 | on_demand_fixed | https://www.hyperstack.cloud/gpu-pricing | none |
| E011 | Verda | T2 | priced | Y | 3.25 | on_demand_fixed | https://verda.com/pricing | none |
| E012 | Voltage Park | T2 | captured-excluded | N |  | starting_floor | https://www.voltagepark.com/; https://www.voltagepark.com/pricing | none |
| E013 | Fluidstack | T2 | captured-excluded | N |  | none_published | https://fluidstack.io/; https://fluidstack.io/pricing; https://fluidstack.io/gpu-clusters; https://fluidstack.io/products; https://fluidstack.io/contact | none |
| E014 | Nscale | T2 | captured-excluded | N |  | sales_or_quote_gated_no_fixed_public_rate | https://www.nscale.com/ ; https://www.nscale.com/product/gpu-nodes ; https://www.nscale.com/gpu-nodes/nvidia-h100 ; https://www.nscale.com/pricing ; https://www.nscale.com/contact | none |
| E015 | IREN | T2 | captured-excluded | N |  | exclusive_pricing_contact_sales | https://iren.com/solutions/gpu-cloud/ai-cloud | none |
| E016 | Applied Digital | T2 | captured-excluded | N |  | no_public_rate_contact_form_404_pricing_route | https://applieddigital.com/ ; https://applieddigital.com/solutions/applied-digital-cloud ; https://applieddigital.com/pricing ; https://applieddigital.com/contact | none |
| E017 | GMI Cloud | T2 | captured-excluded | N |  | starting_floor | https://gmicloud.ai/en/pricing ; https://gmicloud.ai/en | none |
| E018 | CUDO Compute | T2 | captured-excluded | N |  | quote_on_request | https://www.cudocompute.com/ ; https://www.cudocompute.com/pricing ; https://www.cudocompute.com/products/nvidia-h100 ; https://www.cudocompute.com/contact | none |
| E019 | Spheron | T2 | captured-excluded | N |  | marketplace_dynamic | https://www.spheron.network/ ; https://www.spheron.network/pricing/ ; https://www.spheron.network/gpu-rental/h100/ ; https://app.spheron.ai/api/gpu-offers | none |
| E020 | Jarvislabs | T2 | priced | Y | 2.69 | on_demand_fixed | https://jarvislabs.ai/pricing | none |
| E021 | Hyperbolic | T2 | captured-excluded | N |  | no_public_comparable_fixed_rate | https://hyperbolic.ai/ ; https://app.hyperbolic.ai/gpus ; https://docs.hyperbolic.ai/docs/general/billing-payments ; https://app.hyperbolic.ai/gpus/reserved/instant-quote/new?preferredRegion=north_america | none |
| E022 | Atlas Cloud | T2 | priced | Y | 2.95 | on_demand_fixed | https://atlascloud.ai/pricing/gpu | none |
| E023 | RunPod | T3 | priced | Y | 3.29 | on_demand_fixed | https://www.runpod.io/pricing | none |
| E024 | Vast.ai | T3 | captured-excluded | N |  | marketplace_dynamic | https://vast.ai/pricing ; https://vast.ai/hosting | none |
| E025 | Paperspace | T3 | priced | Y | 5.95 | on_demand_fixed | https://paperspace.com/pricing | none |
| E026 | TensorDock | T3 | captured-excluded | N |  | host_marketplace_floor | https://tensordock.com/ ; https://tensordock.com/gpu-h100.html ; https://tensordock.com/cloud-gpus.html | none |
| E027 | Shadeform | T3 | captured-excluded | N |  | aggregator_passthrough | https://shadeform.com/ ; https://shadeform.com/directory/instances ; https://shadeform.com/prices | none |

## Public evidence redaction note

The public evidence package v1.2 crops or redacts E021 and E024 through E027 screenshot frames to remove browser chrome, taskbars, email tabs, bookmarks, and visible signed-in founder-identifying text. This is an evidence-display correction only and does not change any classification.