{"date":"2026-10-26","method":"Every row re-read on the seller's current pricing page (or the same source class it was originally collected from) by WebFetch on 2026-10-26. Status: confirmed = same price read today; changed = different price read today (previous kept as price_previous); unavailable = no longer listed (price cleared, delisted_as_of set); not_rechecked = no readable source today (price and as_of untouched, so staleness stays visible). Quotes are the exact price text read. Per-node prices divided by GPUs per node as noted.","results":[{"i":0,"status":"confirmed","price":6.88,"quote":"On Demand $55.040 ... starting at $55.04 per hour","checked_at":"2026-10-26","note":"confirmed via secondary source https://instances.vantage.sh/aws/ec2/p5.48xlarge; Secondary source (vantage). 55.04/8 GPUs. Default view us-east-1 Linux (per page filter defaults, not explicitly stated).","provider":"AWS"},{"i":1,"status":"confirmed","price":7.912,"quote":"On Demand price of $63.296 per hour","checked_at":"2026-10-26","note":"confirmed via secondary source https://instances.vantage.sh/aws/ec2/p5en.48xlarge; Secondary source (vantage). 63.296/8. Default us-east-1 Linux.","provider":"AWS"},{"i":2,"status":"confirmed","price":14.2416,"quote":"starting at $113.9328 per hour","checked_at":"2026-10-26","note":"confirmed via secondary source https://instances.vantage.sh/aws/ec2/p6-b200.48xlarge; Secondary source (vantage). 113.9328/8. Default us-east-1 Linux.","provider":"AWS"},{"i":3,"status":"confirmed","price":17.802,"quote":"starting at $142.416 per hour","checked_at":"2026-10-26","note":"confirmed via secondary source https://instances.vantage.sh/aws/ec2/p6-b300.48xlarge; Secondary source (vantage). 142.416/8. Default us-east-1 Linux.","provider":"AWS"},{"i":4,"status":"confirmed","price":11.0613,"source_url":"https://cloud.google.com/products/compute/pricing/accelerator-optimized","quote":"a3-highgpu-8g $88.490000119 / 1 hour","checked_at":"2026-10-26","note":"88.49/8. Also matches gcloud-compute.com us-central1 row '88.49'.","provider":"Google Cloud"},{"i":5,"status":"confirmed","price":11.6172,"source_url":"https://gcloud-compute.com/a3-megagpu-8g.html","quote":"us-east4 standard price per hour 92.9376","checked_at":"2026-10-26","note":"92.9376/8, us-east4 (us-central1 not offered). CAUTION: official GCP page default table shows $93.400712807/hr = 11.6751/GPU (region unstated).","provider":"Google Cloud"},{"i":6,"status":"changed","price":10.6009,"source_url":"https://cloud.google.com/products/compute/pricing/accelerator-optimized","quote":"a3-ultragpu-8g Price (USD) $84.806908493 / 1 hour","checked_at":"2026-10-26","note":"Previously no price; new entry. 84.806908493/8. Region of table not stated (Iowa likely default). gcloud-compute page failed to load.","provider":"Google Cloud"},{"i":7,"status":"confirmed","price":null,"source_url":"https://cloud.google.com/products/compute/pricing/accelerator-optimized","quote":"a4-highgpu-8g Price (USD) N/A","checked_at":"2026-10-26","note":"No on-demand price published. Page lists DWS Flex-start $64.44/hr (8.055/GPU), Calendar $90.22/hr (11.2775/GPU), Spot $39.6336/hr (4.9542/GPU).","provider":"Google Cloud"},{"i":8,"status":"confirmed","price":12.29,"source_url":"https://prices.azure.com/api/retail/prices?$filter=armRegionName eq 'eastus' and armSkuName eq 'Standard_ND96isr_H100_v5'","quote":"NDsr H100 v5 Series ND96isrH100v5 Consumption 98.32 1 Hour","checked_at":"2026-10-26","note":"Linux eastus, 98.32/8.","provider":"Microsoft Azure"},{"i":9,"status":"confirmed","price":10.6,"source_url":"https://prices.azure.com/api/retail/prices?$filter=armSkuName eq 'Standard_ND96isr_H200_v5' and armRegionName eq 'eastus2'","quote":"NDsrH200v5 Series ND96isrH200v5 Consumption 84.8 1 Hour","checked_at":"2026-10-26","note":"84.8/8. Not offered in eastus (0 items); eastus2 and westus3 both 84.8 Linux.","provider":"Microsoft Azure"},{"i":10,"status":"confirmed","price":27.04,"source_url":"https://prices.azure.com/api/retail/prices?$filter=armSkuName eq 'Standard_ND128isr_NDR_GB200_v6' and armRegionName eq 'eastus'","quote":"NDsrGB200NDRv6 Series ND128isrNDRGB200v6 Consumption 108.16 1 Hour","checked_at":"2026-10-26","note":"Per VM (4 GPUs): 108.16/4.","provider":"Microsoft Azure"},{"i":11,"status":"confirmed","price":10,"source_url":"https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/?partNumber=B98415","quote":"Compute - GPU - H100, GPU Per Hour, USD 10, PAY_AS_YOU_GO","checked_at":"2026-10-26","note":"","provider":"Oracle Cloud"},{"i":12,"status":"confirmed","price":10,"source_url":"https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/?partNumber=B110519","quote":"OCI - Compute - GPU - H200, GPU Per Hour, USD 10","checked_at":"2026-10-26","note":"","provider":"Oracle Cloud"},{"i":13,"status":"confirmed","price":14,"source_url":"https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/?partNumber=B110978","quote":"OCI - Compute - GPU - B200, GPU Per Hour, USD 14","checked_at":"2026-10-26","note":"","provider":"Oracle Cloud"},{"i":14,"status":"confirmed","price":16,"source_url":"https://apexapps.oracle.com/pls/apex/cetools/api/v1/products/?partNumber=B110979","quote":"OCI - Compute - GPU - GB200, GPU Per Hour, USD 16","checked_at":"2026-10-26","note":"Unit is per GPU-hour.","provider":"Oracle Cloud"},{"i":15,"status":"changed","price":7.99,"source_url":"https://www.runpod.io/pricing","quote":"B200 180 GB VRAM 283 GB RAM 28 vCPUs $7.99/hr","checked_at":"2026-10-26","note":"Page shows one price per row and does not label Secure vs Community; usagepricing.com rate card (RunPod 2026-10-15 Secure Cloud hike) corroborates Secure Cloud B200 $6.79 -> $7.99.","provider":"RunPod"},{"i":16,"status":"changed","price":5.29,"source_url":"https://www.runpod.io/pricing","quote":"H200 141 GB VRAM 276 GB RAM 24 vCPUs $5.29/hr","checked_at":"2026-10-26","note":"Tier unlabeled on page; usagepricing.com corroborates Secure Cloud H200 $4.59 -> $5.29 (2026-10-15).","provider":"RunPod"},{"i":17,"status":"changed","price":3.99,"source_url":"https://www.runpod.io/pricing","quote":"H100 SXM 80 GB VRAM 125 GB RAM 20 vCPUs $3.99/hr","checked_at":"2026-10-26","note":"Tier unlabeled on page; usagepricing.com corroborates Secure Cloud H100 SXM $3.49 -> $3.99 (2026-10-15).","provider":"RunPod"},{"i":18,"status":"confirmed","price":2.89,"source_url":"https://www.runpod.io/pricing","quote":"H100 PCIe 80 GB VRAM 188 GB RAM 16 vCPUs $2.89/hr","checked_at":"2026-10-26","note":"Tier unlabeled on page; usagepricing.com reports H100 PCIe unchanged in the 2026-10-15 Secure Cloud hike.","provider":"RunPod"},{"i":19,"status":"changed","price":1.79,"source_url":"https://www.runpod.io/pricing","quote":"A100 SXM 80 GB VRAM 125 GB RAM 16 vCPUs $1.79/hr","checked_at":"2026-10-26","note":"Tier unlabeled on page; usagepricing.com corroborates Secure Cloud A100 SXM $1.59 -> $1.79 (2026-10-15).","provider":"RunPod"},{"i":20,"status":"changed","price":0.89,"source_url":"https://www.runpod.io/pricing","quote":"RTX 4090 24 GB VRAM 41 GB RAM 6 vCPUs $0.89/hr","checked_at":"2026-10-26","note":"Tier unlabeled on page; usagepricing.com corroborates Secure Cloud RTX 4090 $0.74 -> $0.89 (2026-10-15).","provider":"RunPod"},{"i":21,"status":"confirmed","price":2.25,"source_url":"https://www.tensordock.com/gpu-h100.html","quote":"starting at just $2.25/hour depending on CPU/RAM resources allocated","checked_at":"2026-10-26","note":"Starting-at; table shows $2.25/hr*, 8x minimum. Reserved: monthly $2.00, annual $1.90, 3-yr $1.50.","provider":"TensorDock"},{"i":22,"status":"confirmed","price":1.91,"source_url":"https://www.tensordock.com/gpu-h100.html","quote":"$1.91/hr** (spot minimum bid)","checked_at":"2026-10-26","note":"Table spot figure is a minimum bid that fluctuates; page body separately says $1.30/hour as spot - internal inconsistency on page.","provider":"TensorDock"},{"i":23,"status":"confirmed","price":1.99,"source_url":"https://www.tensordock.com/benchmarks.html","quote":"H100 PCIE 80GB ($1.99)","checked_at":"2026-10-26","note":"Benchmark page label; no explicit time unit stated. H100 product page lists no PCIe pricing.","provider":"TensorDock"},{"i":24,"status":"confirmed","price":1.65,"source_url":"https://www.tensordock.com/benchmarks.html","quote":"A100 SXM4 80GB ($1.65)","checked_at":"2026-10-26","note":"Benchmark page, no time unit stated. Conflict: tensordock.com homepage now says A100 SXM4 'From $1.80/hr' - consider switching to 1.80.","provider":"TensorDock"},{"i":25,"status":"confirmed","price":0.35,"source_url":"https://tensordock.com","quote":"RTX 4090: From $0.35/hr","checked_at":"2026-10-26","note":"Starting-at. Benchmarks page labels RTX 4090 at $0.41/$0.43 (no unit).","provider":"TensorDock"},{"i":26,"status":"confirmed","price":2.43,"source_url":"https://primeintellect.ai/","quote":"H100 $2.43/HR","checked_at":"2026-10-26","note":"Aggregated on-demand, illustrative.","provider":"Prime Intellect"},{"i":27,"status":"confirmed","price":0.94,"source_url":"https://primeintellect.ai/","quote":"Spot 0.94/HR","checked_at":"2026-10-26","note":"H100 spot, illustrative.","provider":"Prime Intellect"},{"i":28,"status":"confirmed","price":3.14,"source_url":"https://primeintellect.ai/","quote":"H200 141 GB VRAM ... $3.14/HR","checked_at":"2026-10-26","note":"Page also shows H200-labeled 80GB-VRAM listings at $1.99/$1.80/$1.23/$0.47 (likely mislabeled; ignored).","provider":"Prime Intellect"},{"i":29,"status":"confirmed","price":3.49,"source_url":"https://primeintellect.ai/","quote":"B200 192 GB VRAM $3.49/hr","checked_at":"2026-10-26","note":"","provider":"Prime Intellect"},{"i":30,"status":"confirmed","price":4.99,"source_url":"https://primeintellect.ai/","quote":"B300 288 GB VRAM $4.99/HR","checked_at":"2026-10-26","note":"Reserved calculator shows B300 SXM6 x512 at $5.00/HR/GPU.","provider":"Prime Intellect"},{"i":31,"status":"confirmed","price":6.0,"source_url":"https://www.hyperstack.cloud/gpu-pricing","quote":"B200 $6.00 per hour (on-demand)","checked_at":"2026-10-26","note":"","provider":"Hyperstack"},{"i":32,"status":"confirmed","price":3.99,"source_url":"https://www.hyperstack.cloud/gpu-pricing","quote":"H200 SXM $3.99 per hour (on-demand)","checked_at":"2026-10-26","note":"","provider":"Hyperstack"},{"i":33,"status":"confirmed","price":3.2,"source_url":"https://www.hyperstack.cloud/gpu-pricing","quote":"H100 SXM $3.20 per hour (on-demand)","checked_at":"2026-10-26","note":"","provider":"Hyperstack"},{"i":34,"status":"confirmed","price":2.72,"source_url":"https://www.hyperstack.cloud/gpu-pricing","quote":"H100 SXM reservation starting from $2.72","checked_at":"2026-10-26","note":"Reserved, starting-from.","provider":"Hyperstack"},{"i":35,"status":"confirmed","price":2.5,"source_url":"https://www.hyperstack.cloud/gpu-pricing","quote":"NVIDIA H100 $2.50 (on-demand)","checked_at":"2026-10-26","note":"Page labels it 'NVIDIA H100' without PCIe; matched via 80GB and PCIe reservation link.","provider":"Hyperstack"},{"i":36,"status":"confirmed","price":1.35,"source_url":"https://www.hyperstack.cloud/gpu-pricing","quote":"NVIDIA A100 $1.35 (on-demand)","checked_at":"2026-10-26","note":"Labeled 'NVIDIA A100' 80GB, matched to PCIe via reservation link. A100 SXM 80GB on-demand is $1.60.","provider":"Hyperstack"},{"i":37,"status":"changed","price":10.91,"source_url":"https://verda.com/pricing","quote":"1x B300 SXM6 268GB | 30 | 255 GB | 268 GB | $10.91/h | $5.455/h","checked_at":"2026-10-26","note":"On-demand 1x instance = per GPU. Was 7.5 (+45%). $5.455/h is the spot column.","provider":"DataCrunch"},{"i":38,"status":"changed","price":8.523,"source_url":"https://verda.com/pricing","quote":"1x B200 SXM6 180GB | 30 | 170 GB | 180 GB | $8.523/h | $4.262/h","checked_at":"2026-10-26","note":"On-demand 1x instance. Instant Clusters table shows $8.525/h per GPU for B200. Was 6.11.","provider":"DataCrunch"},{"i":39,"status":"changed","price":6.006,"source_url":"https://verda.com/pricing","quote":"1x H200 SXM5 141GB | 44 | 170 GB | 141 GB | $6.006/h | $3.003/h","checked_at":"2026-10-26","note":"On-demand 1x instance. Was 4.00.","provider":"DataCrunch"},{"i":40,"status":"changed","price":4.606,"source_url":"https://verda.com/pricing","quote":"1x H100 SXM5 80GB | 30 | 120 GB | 80 GB | $4.606/h | $2.303/h","checked_at":"2026-10-26","note":"On-demand 1x instance (30- and 32-CPU variants same price). Was 3.25.","provider":"DataCrunch"},{"i":41,"status":"changed","price":2.16,"source_url":"https://verda.com/pricing","quote":"1x A100 SXM4 80GB | 22 | 120 GB | 80 GB | $2.160/h | $1.080/h","checked_at":"2026-10-26","note":"On-demand 1x instance. Was 1.79.","provider":"DataCrunch"},{"i":42,"status":"changed","price":1.827,"source_url":"https://verda.com/pricing","quote":"1x L40S 48GB | 20 | 60 GB | 48 GB | $1.827/h | $0.9135/h","checked_at":"2026-10-26","note":"On-demand 1x instance. Was 1.37.","provider":"DataCrunch"},{"i":43,"status":"changed","price":3.57,"source_url":"https://gpuperhour.com/providers/scaleway","quote":"H100 SXM5 80GB $3.57/hr","checked_at":"2026-10-26","note":"Official scaleway.com/en/pricing/gpu/ renders no H100-SXM price (JS-loaded). Aggregator USD value, likely EUR-converted; was 3.61. Drift may be FX, not a list-price change.","provider":"Scaleway","source_class":"aggregator"},{"i":44,"status":"changed","price":3.21,"source_url":"https://gpuperhour.com/providers/scaleway","quote":"H100 PCIe 80GB $3.21/hr","checked_at":"2026-10-26","note":"Aggregator lists 'H100 PCIe' (= H100-1-80G). Official page shows no price; scaleway.com/en/h100/ says 'starting at €2.73/hour'. Was 3.34; likely FX drift.","provider":"Scaleway","source_class":"aggregator"},{"i":45,"status":"changed","price":1.65,"source_url":"https://gpuperhour.com/providers/scaleway","quote":"L40S 48GB $1.65/hr","checked_at":"2026-10-26","note":"Official page shows no L40S price. Aggregator USD value; was 1.68; likely FX drift.","provider":"Scaleway","source_class":"aggregator"},{"i":46,"status":"changed","price":0.88,"source_url":"https://gpuperhour.com/providers/scaleway","quote":"$0.88/GPU/hr","checked_at":"2026-10-26","note":"Official scaleway.com/en/pricing/gpu/ shows L4-1-24G '€0.79 /HOUR' (EUR, excl. tax). USD aggregator 0.88 vs old 0.87: FX drift, EUR list price unverified vs prior.","provider":"Scaleway","source_class":"aggregator"},{"i":47,"status":"not_rechecked","price":null,"source_url":"https://gpuperhour.com/providers/scaleway","quote":"B300 SXM6 $8.39/hr (model list) vs $11.02/hr (breakdown text)","checked_at":"2026-10-26","note":"Official page shows no B300 price; aggregator contradicts itself (8.39 vs 11.02). Not usable. Old 8.74 unverified.","provider":"Scaleway"},{"i":48,"status":"confirmed","price":6.195,"source_url":"https://www.ovhcloud.com/en/public-cloud/gpu/","quote":"H200 From $49.56 /hour *Estimated pricing for 8 GPUs","checked_at":"2026-10-26","note":"Per-instance 8 GPUs: 49.56/8 = 6.195/GPU-hr, rounds to old 6.20. 'From' price, labeled estimate.","provider":"OVHcloud"},{"i":49,"status":"confirmed","price":2.99,"source_url":"https://www.ovhcloud.com/en/public-cloud/gpu/","quote":"H100 From $2.99 /hour","checked_at":"2026-10-26","note":"'From' price; GPU count per instance not stated on page (assumed 1-GPU entry tier).","provider":"OVHcloud"},{"i":50,"status":"confirmed","price":1.8,"source_url":"https://www.ovhcloud.com/en/public-cloud/gpu/","quote":"NVIDIA L40S From $1.80 /hour","checked_at":"2026-10-26","note":"'From' price; GPU count not stated.","provider":"OVHcloud"},{"i":51,"status":"confirmed","price":1.0,"source_url":"https://www.ovhcloud.com/en/public-cloud/gpu/","quote":"Nvidia L4 From $1 /hour","checked_at":"2026-10-26","note":"'From' price; GPU count not stated.","provider":"OVHcloud"},{"i":52,"status":"confirmed","price":8.19,"source_url":"https://www.together.ai/gpu-clusters","quote":"HGX B200 $8.19/hr per GPU","checked_at":"2026-10-26","note":"On-demand; also $8.19 on together.ai/pricing.","provider":"Together AI"},{"i":53,"status":"confirmed","price":6.79,"source_url":"https://www.together.ai/gpu-clusters","quote":"Starting at $6.79/hr per GPU","checked_at":"2026-10-26","note":"Reserved floor; /pricing shows this as 91-180 day tier.","provider":"Together AI"},{"i":54,"status":"confirmed","price":5.99,"source_url":"https://www.together.ai/gpu-clusters","quote":"H200 $5.99/hr per GPU","checked_at":"2026-10-26","note":"On-demand.","provider":"Together AI"},{"i":55,"status":"confirmed","price":3.99,"source_url":"https://www.together.ai/gpu-clusters","quote":"Starting at $3.99/hr per GPU","checked_at":"2026-10-26","note":"Reserved floor; 91-180 day tier on /pricing.","provider":"Together AI"},{"i":56,"status":"confirmed","price":3.99,"source_url":"https://www.together.ai/gpu-clusters","quote":"H100 $3.99/hr per GPU","checked_at":"2026-10-26","note":"On-demand.","provider":"Together AI"},{"i":57,"status":"confirmed","price":3.19,"source_url":"https://www.together.ai/gpu-clusters","quote":"Starting at $3.19/hr per GPU","checked_at":"2026-10-26","note":"Reserved floor; 91-180 day tier on /pricing.","provider":"Together AI"},{"i":58,"status":"changed","price":5.97,"source_url":"https://aws.amazon.com/ec2/capacityblocks/pricing/","quote":"p5.48xlarge US East (N. Virginia) $47.757 USD ($5.970 USD) 8 x H100","checked_at":"2026-10-26","note":"Up ~15% from 5.191. Same rate in all listed regions. Page: next update Jan 2027.","provider":"AWS"},{"i":59,"status":"changed","price":6.866,"source_url":"https://aws.amazon.com/ec2/capacityblocks/pricing/","quote":"p5e.48xlarge US East (Ohio) $54.924 USD ($6.866 USD) 8 x H200","checked_at":"2026-10-26","note":"Up ~15% from 5.97. Not offered in N. Virginia; Ohio/Oregon/etc. all 6.866.","provider":"AWS"},{"i":60,"status":"changed","price":7.895,"source_url":"https://aws.amazon.com/ec2/capacityblocks/pricing/","quote":"p5en.48xlarge US East (N. Virginia) $63.158 USD ($7.895 USD)","checked_at":"2026-10-26","note":"Up ~15% from 6.865.","provider":"AWS"},{"i":61,"status":"changed","price":14.208,"source_url":"https://aws.amazon.com/ec2/capacityblocks/pricing/","quote":"p6-b200.48xlarge US East (N. Virginia) $113.666 USD ($14.208 USD)","checked_at":"2026-10-26","note":"Up ~15% from 12.355. GovCloud 14.801.","provider":"AWS"},{"i":62,"status":"confirmed","price":10.582,"source_url":"https://aws.amazon.com/ec2/capacityblocks/pricing/","quote":"u-p6e-gb200x72 US East (Dallas) Local Zone $761.904 ($10.582 USD) per UltraServer","checked_at":"2026-10-26","note":"$761.904 per UltraServer-hr / 72 GPUs. Dallas Local Zone only. p6-b300 capacity block also listed at $16.146/GPU (not in registry).","provider":"AWS"},{"i":63,"status":"not_rechecked","price":null,"source_url":"https://aws.amazon.com/savingsplans/compute-pricing/","quote":"","checked_at":"2026-10-26","note":"p5 SP 1yr. Rates are JS-rendered; no numbers in fetched text. Vantage shows no SP rates.","provider":"AWS"},{"i":64,"status":"not_rechecked","price":null,"source_url":"https://aws.amazon.com/savingsplans/compute-pricing/","quote":"","checked_at":"2026-10-26","note":"p5 SP 3yr. Not fetchable. FYI vantage p5 3-yr RESERVED (not SP) $23.777/hr = 2.9721/GPU.","provider":"AWS"},{"i":65,"status":"not_rechecked","price":null,"source_url":"https://aws.amazon.com/savingsplans/compute-pricing/","quote":"","checked_at":"2026-10-26","note":"p5en SP 1yr. Not fetchable.","provider":"AWS"},{"i":66,"status":"not_rechecked","price":null,"source_url":"https://aws.amazon.com/savingsplans/compute-pricing/","quote":"","checked_at":"2026-10-26","note":"p5en SP 3yr. Not fetchable.","provider":"AWS"},{"i":67,"status":"not_rechecked","price":null,"source_url":"https://aws.amazon.com/savingsplans/compute-pricing/","quote":"","checked_at":"2026-10-26","note":"p6-b200 SP 1yr. Not fetchable.","provider":"AWS"},{"i":68,"status":"not_rechecked","price":null,"source_url":"https://aws.amazon.com/savingsplans/compute-pricing/","quote":"","checked_at":"2026-10-26","note":"p6-b200 SP 3yr. Not fetchable.","provider":"AWS"},{"i":69,"status":"confirmed","price":7.8656,"source_url":"https://prices.azure.com/api/retail/prices?$filter=armRegionName eq 'eastus' and armSkuName eq 'Standard_ND96isr_H100_v5'","quote":"Reservation 1 Year 551221.0","checked_at":"2026-10-26","note":"Term total: 551221/8760/8 = 7.86560","provider":"Microsoft Azure"},{"i":70,"status":"confirmed","price":5.3953,"source_url":"https://prices.azure.com/api/retail/prices?$filter=armRegionName eq 'eastus' and armSkuName eq 'Standard_ND96isr_H100_v5'","quote":"Reservation 3 Years 1134310.0","checked_at":"2026-10-26","note":"1134310/26280/8 = 5.39531","provider":"Microsoft Azure"},{"i":71,"status":"confirmed","price":5.8152,"source_url":"https://prices.azure.com/api/retail/prices?$filter=armSkuName eq 'Standard_ND96isr_H200_v5' and armRegionName eq 'eastus2'","quote":"Reservation 1 Year 407526.0 effective 2026-08-01","checked_at":"2026-10-26","note":"407526/8760/8 = 5.81515. eastus2.","provider":"Microsoft Azure"},{"i":72,"status":"confirmed","price":5.2777,"source_url":"https://prices.azure.com/api/retail/prices?$filter=armSkuName eq 'Standard_ND96isr_H200_v5' and armRegionName eq 'eastus2'","quote":"Reservation 3 Years 1109592.0 effective 2026-08-01","checked_at":"2026-10-26","note":"1109592/26280/8 = 5.27774. eastus2.","provider":"Microsoft Azure"},{"i":73,"status":"confirmed","price":17.3056,"source_url":"https://prices.azure.com/api/retail/prices?$filter=armSkuName eq 'Standard_ND128isr_NDR_GB200_v6' and armRegionName eq 'eastus'","quote":"Reservation 1 Year 606388.0","checked_at":"2026-10-26","note":"606388/8760/4 GPUs = 17.3056","provider":"Microsoft Azure"},{"i":74,"status":"confirmed","price":11.8976,"source_url":"https://prices.azure.com/api/retail/prices?$filter=armSkuName eq 'Standard_ND128isr_NDR_GB200_v6' and armRegionName eq 'eastus'","quote":"Reservation 3 Years 1250676.0","checked_at":"2026-10-26","note":"1250676/26280/4 = 11.8976","provider":"Microsoft Azure"},{"i":75,"status":"confirmed","price":7.673,"source_url":"https://cloud.google.com/products/compute/pricing/accelerator-optimized","quote":"a3-highgpu-8g 1-yr CUD $61.383674231 / hr","checked_at":"2026-10-26","note":"/8 = 7.67296","provider":"Google Cloud"},{"i":76,"status":"confirmed","price":4.858,"source_url":"https://cloud.google.com/products/compute/pricing/accelerator-optimized","quote":"a3-highgpu-8g 3-yr CUD $38.864383195 / hr","checked_at":"2026-10-26","note":"/8 = 4.85805","provider":"Google Cloud"},{"i":77,"status":"confirmed","price":8.0258,"source_url":"https://cloud.google.com/products/compute/pricing/accelerator-optimized","quote":"a3-megagpu-8g CUD 1 Year $64.206558727 / 1 hour","checked_at":"2026-10-26","note":"/8 = 8.02582","provider":"Google Cloud"},{"i":78,"status":"confirmed","price":5.0813,"source_url":"https://cloud.google.com/products/compute/pricing/accelerator-optimized","quote":"a3-megagpu-8g CUD 3 Year $40.650103355 / 1 hour","checked_at":"2026-10-26","note":"/8 = 5.08126","provider":"Google Cloud"},{"i":79,"status":"confirmed","price":7.309,"source_url":"https://cloud.google.com/products/compute/pricing/accelerator-optimized","quote":"a3-ultragpu-8g CUD 1 Year $58.471933151 / 1 hour","checked_at":"2026-10-26","note":"/8 = 7.30899","provider":"Google Cloud"},{"i":80,"status":"confirmed","price":4.6511,"source_url":"https://cloud.google.com/products/compute/pricing/accelerator-optimized","quote":"a3-ultragpu-8g CUD 3 Year $37.208420822 / 1 hour","checked_at":"2026-10-26","note":"/8 = 4.65105","provider":"Google Cloud"},{"i":81,"status":"confirmed","price":11.1159,"source_url":"https://cloud.google.com/products/compute/pricing/accelerator-optimized","quote":"a4-highgpu-8g CUD 1 Year $88.9272 / 1 hour","checked_at":"2026-10-26","note":"/8 = 11.1159. Row cells extracted misaligned once, re-read with header mapping.","provider":"Google Cloud"},{"i":82,"status":"confirmed","price":7.0884,"source_url":"https://cloud.google.com/products/compute/pricing/accelerator-optimized","quote":"a4-highgpu-8g CUD 3 Year $56.7072 / 1 hour","checked_at":"2026-10-26","note":"/8 = 7.0884","provider":"Google Cloud"},{"i":83,"status":"not_rechecked","price":null,"source_url":"https://lambda.ai/service/gpu-cloud/1-click-clusters","quote":"","checked_at":"2026-10-26","note":"Page has an H100 tab but its prices are not in the fetched content (likely client-side tab). Only B200 prices readable. Needs a browser check; do not treat as removed.","provider":"Lambda"},{"i":84,"status":"not_rechecked","price":null,"source_url":"https://lambda.ai/service/gpu-cloud/1-click-clusters","quote":"","checked_at":"2026-10-26","note":"Same as 83: H100 tab prices not readable via fetch.","provider":"Lambda"},{"i":85,"status":"confirmed","price":9.36,"source_url":"https://lambda.ai/service/gpu-cloud/1-click-clusters","quote":"NVIDIA HGX B200, 64 GPUs, 2 weeks - 1 year, $9.36","checked_at":"2026-10-26","note":"64-GPU tier, 2 weeks-1 year term: $9.36/GPU-hr, plus tax. 1 year+ tier shows no price ('Talk to our team').","provider":"Lambda"},{"i":86,"status":"confirmed","price":8.87,"source_url":"https://lambda.ai/service/gpu-cloud/1-click-clusters","quote":"NVIDIA HGX B200, 256+ GPUs, 2 weeks - 1 year, $8.87","checked_at":"2026-10-26","note":"256+ GPU tier, 2 weeks-1 year term.","provider":"Lambda"},{"i":87,"status":"confirmed","price":2.79,"source_url":"https://www.hyperstack.cloud/gpu-pricing","quote":"H200 SXM reservation starting from $2.79","checked_at":"2026-10-26","note":"Starting-from.","provider":"Hyperstack"},{"i":88,"status":"confirmed","price":1.75,"source_url":"https://www.hyperstack.cloud/gpu-pricing","quote":"NVIDIA H100 reservation starting from $1.75","checked_at":"2026-10-26","note":"PCIe inferred from reservation link; starting-from.","provider":"Hyperstack"},{"i":89,"status":"confirmed","price":1.36,"source_url":"https://www.hyperstack.cloud/gpu-pricing","quote":"A100 SXM 80GB reservation starting from $1.36","checked_at":"2026-10-26","note":"Starting-from. A100 PCIe reserved is $0.95 (not in registry).","provider":"Hyperstack"},{"i":90,"status":"confirmed","price":5.1,"source_url":"https://www.hyperstack.cloud/gpu-pricing","quote":"B200 reservation starting from $5.10","checked_at":"2026-10-26","note":"Starting-from.","provider":"Hyperstack"},{"i":91,"status":"unavailable","price":null,"source_url":"https://verda.com/pricing","quote":"1 year | 8% off on-demand","checked_at":"2026-10-26","note":"No explicit $ long-term price published; only % discount off on-demand plus 'Get a custom quote'. Not computing a derived price.","provider":"DataCrunch"},{"i":92,"status":"unavailable","price":null,"source_url":"https://verda.com/pricing","quote":"2 years | 25% off on-demand","checked_at":"2026-10-26","note":"No explicit $ long-term price published; only % discount off on-demand.","provider":"DataCrunch"},{"i":93,"status":"unavailable","price":null,"source_url":"https://verda.com/pricing","quote":"1 year | 8% off on-demand","checked_at":"2026-10-26","note":"No explicit $ long-term price published; only % discount off on-demand.","provider":"DataCrunch"},{"i":94,"status":"unavailable","price":null,"source_url":"https://verda.com/pricing","quote":"2 years | 25% off on-demand","checked_at":"2026-10-26","note":"No explicit $ long-term price published; only % discount off on-demand.","provider":"DataCrunch"},{"i":95,"status":"unavailable","price":null,"source_url":"https://verda.com/pricing","quote":"1 year | 8% off on-demand","checked_at":"2026-10-26","note":"No explicit $ long-term price published; only % discount off on-demand.","provider":"DataCrunch"},{"i":96,"status":"unavailable","price":null,"source_url":"https://verda.com/pricing","quote":"2 years | 25% off on-demand","checked_at":"2026-10-26","note":"No explicit $ long-term price published; only % discount off on-demand.","provider":"DataCrunch"},{"i":97,"status":"unavailable","price":null,"source_url":"https://verda.com/pricing","quote":"1 year | 8% off on-demand","checked_at":"2026-10-26","note":"No explicit $ long-term price published; only % discount off on-demand.","provider":"DataCrunch"},{"i":98,"status":"unavailable","price":null,"source_url":"https://verda.com/pricing","quote":"2 years | 25% off on-demand","checked_at":"2026-10-26","note":"No explicit $ long-term price published; only % discount off on-demand.","provider":"DataCrunch"},{"i":99,"status":"confirmed","price":3.69,"source_url":"https://www.together.ai/pricing","quote":"H100 Reserved 3-30 days $3.69","checked_at":"2026-10-26","note":"Price unchanged but tier now labeled 3-30 days (was 7-30 days). Rename SKU.","provider":"Together AI","sku_rename":"Reserved H100, 3-30 days (3-30 days)"},{"i":100,"status":"confirmed","price":3.45,"source_url":"https://www.together.ai/pricing","quote":"H100 Reserved 31-90 days $3.45","checked_at":"2026-10-26","note":"Per GPU-hour.","provider":"Together AI"},{"i":101,"status":"confirmed","price":4.99,"source_url":"https://www.together.ai/pricing","quote":"H200 Reserved 3-30 days $4.99","checked_at":"2026-10-26","note":"Tier now labeled 3-30 days (was 7-30 days).","provider":"Together AI","sku_rename":"Reserved H200, 3-30 days (3-30 days)"},{"i":102,"status":"confirmed","price":4.15,"source_url":"https://www.together.ai/pricing","quote":"H200 Reserved 31-90 days $4.15","checked_at":"2026-10-26","note":"Per GPU-hour.","provider":"Together AI"},{"i":103,"status":"confirmed","price":7.99,"source_url":"https://www.together.ai/pricing","quote":"B200 Reserved 3-30 days $7.99","checked_at":"2026-10-26","note":"Tier now labeled 3-30 days (was 7-30 days).","provider":"Together AI","sku_rename":"Reserved B200, 3-30 days (3-30 days)"},{"i":104,"status":"confirmed","price":7.79,"source_url":"https://www.together.ai/pricing","quote":"B200 Reserved 31-90 days $7.79","checked_at":"2026-10-26","note":"Per GPU-hour.","provider":"Together AI"},{"i":105,"status":"confirmed","price":10.5,"source_url":"https://www.coreweave.com/pricing","quote":"NVIDIA GB200 NVL72, GPU count 4, On-Demand $42.00 / Hour","checked_at":"2026-10-26","note":"Per-instance price $42.00 / 4 GPUs = $10.50/GPU-hr. Same in North America and Europe. Page also shows '$10.50 / Hour' in the GB200 detail block, labeled inconsistently (single CPU vs single GPU price).","provider":"CoreWeave"},{"i":106,"status":"confirmed","price":8.6,"source_url":"https://www.coreweave.com/pricing","quote":"NVIDIA HGX B200, GPU count 8, On-Demand $68.80 / Hour","checked_at":"2026-10-26","note":"Per-instance $68.80 / 8 GPUs = $8.60/GPU-hr. Spot now also listed: $34.11/hr NA (= $4.26/GPU-hr).","provider":"CoreWeave"},{"i":107,"status":"confirmed","price":6.31,"source_url":"https://www.coreweave.com/pricing","quote":"NVIDIA HGX H200, GPU count 8, On-Demand $50.44 / Hour","checked_at":"2026-10-26","note":"Per-instance $50.44 / 8 = $6.305/GPU-hr, rounds to 6.31. Page does not show the gd-8xh200ib-i128 SKU string.","provider":"CoreWeave"},{"i":108,"status":"confirmed","price":6.16,"source_url":"https://www.coreweave.com/pricing","quote":"NVIDIA HGX H100, GPU count 8, On-Demand $49.24 / Hour","checked_at":"2026-10-26","note":"Per-instance $49.24 / 8 = $6.155/GPU-hr, rounds to 6.16. Page does not show the gd-8xh100ib-i128 SKU string.","provider":"CoreWeave"},{"i":109,"status":"confirmed","price":2.46,"source_url":"https://www.coreweave.com/pricing","quote":"NVIDIA HGX H100, GPU count 8, Spot $19.71 / Hour (North America)","checked_at":"2026-10-26","note":"Per-instance $19.71 / 8 = $2.464/GPU-hr (NA). Europe spot is $19.51/hr = $2.439/GPU-hr. Spot is variable.","provider":"CoreWeave"},{"i":110,"status":"confirmed","price":2.7,"source_url":"https://www.coreweave.com/pricing","quote":"NVIDIA A100, GPU count 8, On-Demand $21.60 / Hour","checked_at":"2026-10-26","note":"Per-instance $21.60 / 8 = $2.70/GPU-hr.","provider":"CoreWeave"},{"i":111,"status":"confirmed","price":6.69,"source_url":"https://lambda.ai/instances","quote":"NVIDIA B200 SXM6, 180 GB, PRICE/GPU/HR* $6.69","checked_at":"2026-10-26","note":"Page gives one per-GPU rate per GPU type; 8x/4x/2x/1x configs not priced separately. Excludes sales tax/VAT.","provider":"Lambda"},{"i":112,"status":"confirmed","price":3.99,"source_url":"https://lambda.ai/instances","quote":"NVIDIA H100 SXM, 80 GB, PRICE/GPU/HR* $3.99","checked_at":"2026-10-26","note":"Single per-GPU rate; 8x config not priced separately.","provider":"Lambda"},{"i":113,"status":"unavailable","price":null,"source_url":"https://lambda.ai/instances","quote":"Table lists B200 SXM6, H100 SXM, A100 SXM 80GB/40GB, Tesla V100 only","checked_at":"2026-10-26","note":"H100 PCIe no longer appears anywhere on the instances page (checked twice).","provider":"Lambda"},{"i":114,"status":"confirmed","price":2.79,"source_url":"https://lambda.ai/instances","quote":"NVIDIA A100 SXM, 80 GB, PRICE/GPU/HR* $2.79","checked_at":"2026-10-26","note":"Single per-GPU rate; 8x config not priced separately.","provider":"Lambda"},{"i":115,"status":"confirmed","price":9.86,"source_url":"https://lambda.ai/1-click-clusters","quote":"NVIDIA HGX B200, 16 GPUs, 2 weeks - 1 year, $9.86","checked_at":"2026-10-26","note":"$9.86 is the 16-GPU tier (2 weeks-1 year). Same content as /service/gpu-cloud/1-click-clusters; duplicates the tier table used by 85/86. No on-demand pricing on page.","provider":"Lambda"},{"i":116,"status":"not_rechecked","price":null,"source_url":"https://lambda.ai/1-click-clusters","quote":"","checked_at":"2026-10-26","note":"H100 tab exists but prices not present in fetched content. Needs browser check.","provider":"Lambda"},{"i":117,"status":"not_rechecked","price":null,"source_url":"https://www.crusoe.ai/cloud/pricing","quote":"$4.29 / GPU-hr","checked_at":"2026-10-26","note":"Value $4.29 is on the page but the rendered text does not tie prices to rows. If table is read column-wise (12 on-demand cells in GPU order, then 12 spot cells), H200=$4.29, matching old value; this is inference, not verified. Page now lists Vera Rubin/GB300/B300/GB200/B200/H200/H100/A100 SXM/A100 PCIe/L40S/MI355X/MI300X; all spot cells 'Contact sales'. Needs browser check.","provider":"Crusoe"},{"i":118,"status":"not_rechecked","price":null,"source_url":"https://www.crusoe.ai/cloud/pricing","quote":"$3.90 / GPU-hr","checked_at":"2026-10-26","note":"$3.90 present; column-wise reading gives H100=$3.90 (matches old) but mapping not explicit. Needs browser check.","provider":"Crusoe"},{"i":119,"status":"not_rechecked","price":null,"source_url":"https://www.crusoe.ai/cloud/pricing","quote":"$3.45 / GPU-hr","checked_at":"2026-10-26","note":"$3.45 present; column-wise reading gives MI300X=$3.45 (matches old) but mapping not explicit. Needs browser check.","provider":"Crusoe"},{"i":120,"status":"not_rechecked","price":null,"source_url":"https://www.crusoe.ai/cloud/pricing","quote":"$2.30 / GPU-hr","checked_at":"2026-10-26","note":"$2.30 present; column-wise reading gives A100 80GB SXM=$2.30 (matches old). Conflicting stale sources (blog.crusoecloud.com/cloud/pricing and Oct-2025 support FAQ) say A100 80GB SXM $1.95; ignore those as outdated. Needs browser check.","provider":"Crusoe"},{"i":121,"status":"changed","price":8.5,"source_url":"https://nebius.com/prices","quote":"NVIDIA HGX B200 | 20 | 224 | from $0.99 | $8.50","checked_at":"2026-10-26","note":"On-demand per GPU-hr rose from $7.15 to $8.50. Excludes VAT.","provider":"Nebius"},{"i":122,"status":"changed","price":0.99,"source_url":"https://nebius.com/prices","quote":"NVIDIA HGX B200 ... Preemptible, GPU-hour * from $0.99","checked_at":"2026-10-26","note":"Preemptible is now dynamic spot pricing (changes as often as every 15 min); $0.99 is a 'from' floor, not a fixed rate. Footnote: highest rate is one cent below on-demand. Not comparable to old fixed $3.95; registry should flag as floor/variable.","provider":"Nebius","price_qualifier":"variable-floor"},{"i":123,"status":"changed","price":5.4,"source_url":"https://nebius.com/prices","quote":"NVIDIA HGX H200 | 16 | 200 | from $0.79 | $5.40","checked_at":"2026-10-26","note":"On-demand per GPU-hr rose from $4.50 to $5.40.","provider":"Nebius"},{"i":124,"status":"changed","price":0.79,"source_url":"https://nebius.com/prices","quote":"NVIDIA HGX H200 ... Preemptible from $0.79","checked_at":"2026-10-26","note":"Dynamic spot floor ('from'), not fixed rate; ceiling is on-demand minus $0.01. Not comparable to old $2.45.","provider":"Nebius","price_qualifier":"variable-floor"},{"i":125,"status":"changed","price":4.5,"source_url":"https://nebius.com/prices","quote":"NVIDIA HGX H100 | 16 | 200 | from $0.79 | $4.50","checked_at":"2026-10-26","note":"On-demand per GPU-hr rose from $3.85 to $4.50.","provider":"Nebius"},{"i":126,"status":"changed","price":0.79,"source_url":"https://nebius.com/prices","quote":"NVIDIA HGX H100 ... Preemptible from $0.79","checked_at":"2026-10-26","note":"Dynamic spot floor ('from'), not fixed rate; ceiling is on-demand minus $0.01. Not comparable to old $2.15.","provider":"Nebius","price_qualifier":"variable-floor"},{"i":127,"status":"confirmed","price":1.99,"source_url":"https://voltagepark.com/pricing","quote":"Pricing for NVIDIA H100s starts at $1.99/hr for on-demand compute.","checked_at":"2026-10-26","note":"Still a starting-at teaser: config unspecified; On-Demand Ethernet, On-Demand 3200 Gbps InfiniBand and Long-Term Reserve tiers all 'Contact for pricing'. FAQ says 1 GPU rentable for 1 hr at that rate, no contract.","provider":"Voltage Park"},{"i":128,"status":"not_rechecked","price":null,"source_url":"https://vast.ai/pricing","quote":"","checked_at":"2026-10-26","note":"vast.ai/pricing and /pricing/gpu/H200 render no prices to fetcher ('No current offers'). Console bundles API returned inconsistent snapshots: cheapest 8x H200 dph_total seen 22.11 (unverified, rentable:false), 29.48 (verified, rentable:false); rentable-filtered query returned one offer at 43.20 (deverified). Too unstable to record.","provider":"Vast.ai"},{"i":129,"status":"not_rechecked","price":null,"source_url":"https://vast.ai/pricing","quote":"","checked_at":"2026-10-26","note":"Pricing page/GPU page show no prices to fetcher. One API snapshot showed a single 8x H100 PCIE offer at dph_total 26.43 (=3.30/GPU) but filtered re-queries returned unrelated GPUs; not reliable.","provider":"Vast.ai"},{"i":130,"status":"not_rechecked","price":null,"source_url":"https://vast.ai/pricing","quote":"","checked_at":"2026-10-26","note":"No prices on pricing/GPU page. One API snapshot: cheapest 8x A100 SXM4 dph_total 4.268 (=0.53/GPU), unconfirmed rentable status; filtered re-queries broken.","provider":"Vast.ai"},{"i":131,"status":"not_rechecked","price":null,"source_url":"https://vast.ai/pricing","quote":"","checked_at":"2026-10-26","note":"No prices on pricing/GPU page. One API snapshot: cheapest 8x RTX 5090 dph_total 4.271 (=0.53/GPU), unconfirmed rentable status.","provider":"Vast.ai"},{"i":132,"status":"not_rechecked","price":null,"source_url":"https://vast.ai/pricing","quote":"","checked_at":"2026-10-26","note":"No prices on pricing/GPU page. API snapshot: cheapest 8x RTX 4090 at dph_total 0.96 (deverified, rentable:false); next verified offers 2.88-3.10 total (~0.36-0.39/GPU), rentable status unconfirmed.","provider":"Vast.ai"}]}