{"updatedAt":"2026-10-31","totalSkus":43,"providerCount":16,"note":"Prices are verified indicative on-demand list prices normalized by memory and compute bandwidth across neocloud and hyperscaler providers.","offers":[{"id":"vast-rtx3090-single","provider":"Vast.ai","providerType":"neocloud","gpuModel":"NVIDIA GeForce RTX 3090","architecture":"Ampere","gpuCount":1,"vramGb":24,"totalVramGb":24,"interconnect":"PCIe Gen4","onDemandHourly":0.29,"spotHourly":0.14,"fp8Tflops":0,"fp16Tflops":142,"vCpu":16,"systemRamGb":64,"region":"US Central","affiliateUrl":"/go/vast-ai","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0121,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":489.7,"monthlyRunRate730h":211.7},"spotMetrics":{"costPerGbVramHr":0.0058,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":1014.3,"monthlyRunRate730h":102.2}},{"id":"runpod-rtx3090-community","provider":"RunPod","providerType":"neocloud","gpuModel":"NVIDIA GeForce RTX 3090","architecture":"Ampere","gpuCount":1,"vramGb":24,"totalVramGb":24,"interconnect":"PCIe Gen4","onDemandHourly":0.44,"spotHourly":0.29,"fp8Tflops":0,"fp16Tflops":142,"vCpu":16,"systemRamGb":64,"region":"Global Community","affiliateUrl":"/go/runpod","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0183,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":322.7,"monthlyRunRate730h":321.2},"spotMetrics":{"costPerGbVramHr":0.0121,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":489.7,"monthlyRunRate730h":211.7}},{"id":"vast-rtx4080-single","provider":"Vast.ai","providerType":"neocloud","gpuModel":"NVIDIA GeForce RTX 4080","architecture":"Ada Lovelace","gpuCount":1,"vramGb":16,"totalVramGb":16,"interconnect":"PCIe Gen4","onDemandHourly":0.34,"spotHourly":0.22,"fp8Tflops":390,"fp16Tflops":195,"vCpu":16,"systemRamGb":64,"region":"EU West","affiliateUrl":"/go/vast-ai","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0213,"fp8TflopsPerDollarHr":1147.1,"fp16TflopsPerDollarHr":573.5,"monthlyRunRate730h":248.2},"spotMetrics":{"costPerGbVramHr":0.0138,"fp8TflopsPerDollarHr":1772.7,"fp16TflopsPerDollarHr":886.4,"monthlyRunRate730h":160.6}},{"id":"vast-rtx4090-single","provider":"Vast.ai","providerType":"neocloud","gpuModel":"NVIDIA GeForce RTX 4090","architecture":"Ada Lovelace","gpuCount":1,"vramGb":24,"totalVramGb":24,"interconnect":"PCIe Gen4","onDemandHourly":0.37,"spotHourly":0.22,"fp8Tflops":660,"fp16Tflops":330,"vCpu":16,"systemRamGb":64,"region":"US Central","affiliateUrl":"/go/vast-ai","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0154,"fp8TflopsPerDollarHr":1783.8,"fp16TflopsPerDollarHr":891.9,"monthlyRunRate730h":270.1},"spotMetrics":{"costPerGbVramHr":0.0092,"fp8TflopsPerDollarHr":3000,"fp16TflopsPerDollarHr":1500,"monthlyRunRate730h":160.6}},{"id":"runpod-rtx4090-single","provider":"RunPod","providerType":"neocloud","gpuModel":"NVIDIA GeForce RTX 4090","architecture":"Ada Lovelace","gpuCount":1,"vramGb":24,"totalVramGb":24,"interconnect":"PCIe Gen4","onDemandHourly":0.74,"spotHourly":0.44,"fp8Tflops":660,"fp16Tflops":330,"vCpu":16,"systemRamGb":64,"region":"US East (Secure)","affiliateUrl":"/go/runpod","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0308,"fp8TflopsPerDollarHr":891.9,"fp16TflopsPerDollarHr":445.9,"monthlyRunRate730h":540.2},"spotMetrics":{"costPerGbVramHr":0.0183,"fp8TflopsPerDollarHr":1500,"fp16TflopsPerDollarHr":750,"monthlyRunRate730h":321.2}},{"id":"lambda-rtx4090-single","provider":"Lambda Labs","providerType":"neocloud","gpuModel":"NVIDIA GeForce RTX 4090","architecture":"Ada Lovelace","gpuCount":1,"vramGb":24,"totalVramGb":24,"interconnect":"PCIe Gen4","onDemandHourly":1.25,"spotHourly":null,"fp8Tflops":660,"fp16Tflops":330,"vCpu":32,"systemRamGb":128,"region":"US West","affiliateUrl":"/go/lambdalabs","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0521,"fp8TflopsPerDollarHr":528,"fp16TflopsPerDollarHr":264,"monthlyRunRate730h":912.5}},{"id":"hyperstack-rtx5090-single","provider":"Hyperstack","providerType":"neocloud","gpuModel":"NVIDIA GeForce RTX 5090","architecture":"Blackwell","gpuCount":1,"vramGb":32,"totalVramGb":32,"interconnect":"PCIe Gen5","onDemandHourly":1.45,"spotHourly":0.95,"fp8Tflops":1290,"fp16Tflops":645,"vCpu":24,"systemRamGb":128,"region":"EU North","affiliateUrl":"/go/hyperstack","isDirectSponsor":true,"sponsorBadge":"Featured Blackwell Partner","inStock":true,"metrics":{"costPerGbVramHr":0.0453,"fp8TflopsPerDollarHr":889.7,"fp16TflopsPerDollarHr":444.8,"monthlyRunRate730h":1058.5},"spotMetrics":{"costPerGbVramHr":0.0297,"fp8TflopsPerDollarHr":1357.9,"fp16TflopsPerDollarHr":678.9,"monthlyRunRate730h":693.5}},{"id":"runpod-l4-single","provider":"RunPod","providerType":"neocloud","gpuModel":"NVIDIA L4","architecture":"Ada Lovelace","gpuCount":1,"vramGb":24,"totalVramGb":24,"interconnect":"PCIe Gen4","onDemandHourly":0.65,"spotHourly":0.4,"fp8Tflops":485,"fp16Tflops":242,"vCpu":16,"systemRamGb":64,"region":"US Central","affiliateUrl":"/go/runpod","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0271,"fp8TflopsPerDollarHr":746.2,"fp16TflopsPerDollarHr":372.3,"monthlyRunRate730h":474.5},"spotMetrics":{"costPerGbVramHr":0.0167,"fp8TflopsPerDollarHr":1212.5,"fp16TflopsPerDollarHr":605,"monthlyRunRate730h":292}},{"id":"aws-g5-a10g-single","provider":"AWS EC2 (g5.xlarge)","providerType":"hyperscaler","gpuModel":"NVIDIA A10G","architecture":"Ampere","gpuCount":1,"vramGb":24,"totalVramGb":24,"interconnect":"PCIe Gen4","onDemandHourly":1.006,"spotHourly":0.42,"fp8Tflops":0,"fp16Tflops":125,"vCpu":4,"systemRamGb":16,"region":"us-east-1","affiliateUrl":"/go/aws","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0419,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":124.3,"monthlyRunRate730h":734.38},"spotMetrics":{"costPerGbVramHr":0.0175,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":297.6,"monthlyRunRate730h":306.6}},{"id":"genesis-rtx-a6000","provider":"Genesis Cloud","providerType":"neocloud","gpuModel":"NVIDIA RTX A6000","architecture":"Ampere","gpuCount":1,"vramGb":48,"totalVramGb":48,"interconnect":"PCIe Gen4","onDemandHourly":0.55,"spotHourly":0.35,"fp8Tflops":0,"fp16Tflops":155,"vCpu":16,"systemRamGb":96,"region":"Iceland","affiliateUrl":"/go/genesiscloud","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0115,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":281.8,"monthlyRunRate730h":401.5},"spotMetrics":{"costPerGbVramHr":0.0073,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":442.9,"monthlyRunRate730h":255.5}},{"id":"runpod-a40-single","provider":"RunPod","providerType":"neocloud","gpuModel":"NVIDIA A40","architecture":"Ampere","gpuCount":1,"vramGb":48,"totalVramGb":48,"interconnect":"PCIe Gen4","onDemandHourly":0.79,"spotHourly":0.45,"fp8Tflops":0,"fp16Tflops":150,"vCpu":16,"systemRamGb":128,"region":"EU Central","affiliateUrl":"/go/runpod","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0165,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":189.9,"monthlyRunRate730h":576.7},"spotMetrics":{"costPerGbVramHr":0.0094,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":333.3,"monthlyRunRate730h":328.5}},{"id":"runpod-rtx-6000-ada","provider":"RunPod","providerType":"neocloud","gpuModel":"NVIDIA RTX 6000 Ada Generation","architecture":"Ada Lovelace","gpuCount":1,"vramGb":48,"totalVramGb":48,"interconnect":"PCIe Gen4","onDemandHourly":0.99,"spotHourly":0.7,"fp8Tflops":733,"fp16Tflops":366,"vCpu":24,"systemRamGb":128,"region":"US East","affiliateUrl":"/go/runpod","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0206,"fp8TflopsPerDollarHr":740.4,"fp16TflopsPerDollarHr":369.7,"monthlyRunRate730h":722.7},"spotMetrics":{"costPerGbVramHr":0.0146,"fp8TflopsPerDollarHr":1047.1,"fp16TflopsPerDollarHr":522.9,"monthlyRunRate730h":511}},{"id":"lambda-rtx-6000-ada","provider":"Lambda Labs","providerType":"neocloud","gpuModel":"NVIDIA RTX 6000 Ada Generation","architecture":"Ada Lovelace","gpuCount":1,"vramGb":48,"totalVramGb":48,"interconnect":"PCIe Gen4","onDemandHourly":1.1,"spotHourly":null,"fp8Tflops":733,"fp16Tflops":366,"vCpu":32,"systemRamGb":128,"region":"US West","affiliateUrl":"/go/lambdalabs","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0229,"fp8TflopsPerDollarHr":666.4,"fp16TflopsPerDollarHr":332.7,"monthlyRunRate730h":803}},{"id":"vultr-l40s-single","provider":"Vultr","providerType":"neocloud","gpuModel":"NVIDIA L40S","architecture":"Ada Lovelace","gpuCount":1,"vramGb":48,"totalVramGb":48,"interconnect":"PCIe Gen4","onDemandHourly":1.5,"spotHourly":0.98,"fp8Tflops":733,"fp16Tflops":366,"vCpu":16,"systemRamGb":128,"region":"NJ, US","affiliateUrl":"/go/vultr","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0313,"fp8TflopsPerDollarHr":488.7,"fp16TflopsPerDollarHr":244,"monthlyRunRate730h":1095},"spotMetrics":{"costPerGbVramHr":0.0204,"fp8TflopsPerDollarHr":748,"fp16TflopsPerDollarHr":373.5,"monthlyRunRate730h":715.4}},{"id":"runpod-a100-40gb-sxm","provider":"RunPod","providerType":"neocloud","gpuModel":"NVIDIA A100 SXM4 (40GB)","architecture":"Ampere","gpuCount":1,"vramGb":40,"totalVramGb":40,"interconnect":"NVLink 600GB/s","onDemandHourly":1.89,"spotHourly":1.19,"fp8Tflops":0,"fp16Tflops":312,"vCpu":30,"systemRamGb":200,"region":"US East","affiliateUrl":"/go/runpod","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0473,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":165.1,"monthlyRunRate730h":1379.7},"spotMetrics":{"costPerGbVramHr":0.0298,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":262.2,"monthlyRunRate730h":868.7}},{"id":"vast-a100-80gb-sxm","provider":"Vast.ai","providerType":"neocloud","gpuModel":"NVIDIA A100 SXM4 (80GB)","architecture":"Ampere","gpuCount":1,"vramGb":80,"totalVramGb":80,"interconnect":"NVLink 600GB/s","onDemandHourly":2.25,"spotHourly":1.45,"fp8Tflops":0,"fp16Tflops":624,"vCpu":32,"systemRamGb":256,"region":"Global Mesh","affiliateUrl":"/go/vast-ai","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0281,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":277.3,"monthlyRunRate730h":1642.5},"spotMetrics":{"costPerGbVramHr":0.0181,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":430.3,"monthlyRunRate730h":1058.5}},{"id":"lambda-a100-80gb-sxm","provider":"Lambda Labs","providerType":"neocloud","gpuModel":"NVIDIA A100 SXM4 (80GB)","architecture":"Ampere","gpuCount":1,"vramGb":80,"totalVramGb":80,"interconnect":"NVLink 600GB/s","onDemandHourly":2.49,"spotHourly":null,"fp8Tflops":0,"fp16Tflops":624,"vCpu":64,"systemRamGb":256,"region":"US Central","affiliateUrl":"/go/lambdalabs","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0311,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":250.6,"monthlyRunRate730h":1817.7}},{"id":"runpod-h100-pcie","provider":"RunPod","providerType":"neocloud","gpuModel":"NVIDIA H100 PCIe (80GB)","architecture":"Hopper","gpuCount":1,"vramGb":80,"totalVramGb":80,"interconnect":"PCIe Gen5","onDemandHourly":2.69,"spotHourly":1.85,"fp8Tflops":1513,"fp16Tflops":756,"vCpu":32,"systemRamGb":256,"region":"US Central","affiliateUrl":"/go/runpod","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0336,"fp8TflopsPerDollarHr":562.5,"fp16TflopsPerDollarHr":281,"monthlyRunRate730h":1963.7},"spotMetrics":{"costPerGbVramHr":0.0231,"fp8TflopsPerDollarHr":817.8,"fp16TflopsPerDollarHr":408.6,"monthlyRunRate730h":1350.5}},{"id":"vast-h100-80gb-sxm5","provider":"Vast.ai","providerType":"neocloud","gpuModel":"NVIDIA H100 SXM5","architecture":"Hopper","gpuCount":1,"vramGb":80,"totalVramGb":80,"interconnect":"NVLink 900GB/s","onDemandHourly":2.99,"spotHourly":1.95,"fp8Tflops":1979,"fp16Tflops":989,"vCpu":32,"systemRamGb":256,"region":"US East","affiliateUrl":"/go/vast-ai","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0374,"fp8TflopsPerDollarHr":661.9,"fp16TflopsPerDollarHr":330.8,"monthlyRunRate730h":2182.7},"spotMetrics":{"costPerGbVramHr":0.0244,"fp8TflopsPerDollarHr":1014.9,"fp16TflopsPerDollarHr":507.2,"monthlyRunRate730h":1423.5}},{"id":"vultr-h100-nvl","provider":"Vultr","providerType":"neocloud","gpuModel":"NVIDIA H100 NVL (94GB)","architecture":"Hopper","gpuCount":1,"vramGb":94,"totalVramGb":94,"interconnect":"NVLink 600GB/s","onDemandHourly":2.89,"spotHourly":2.1,"fp8Tflops":1670,"fp16Tflops":835,"vCpu":32,"systemRamGb":256,"region":"Silicon Valley, US","affiliateUrl":"/go/vultr","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0307,"fp8TflopsPerDollarHr":577.9,"fp16TflopsPerDollarHr":288.9,"monthlyRunRate730h":2109.7},"spotMetrics":{"costPerGbVramHr":0.0223,"fp8TflopsPerDollarHr":795.2,"fp16TflopsPerDollarHr":397.6,"monthlyRunRate730h":1533}},{"id":"lambda-h100-80gb-sxm5","provider":"Lambda Labs","providerType":"neocloud","gpuModel":"NVIDIA H100 SXM5","architecture":"Hopper","gpuCount":1,"vramGb":80,"totalVramGb":80,"interconnect":"NVLink 900GB/s","onDemandHourly":3.29,"spotHourly":null,"fp8Tflops":1979,"fp16Tflops":989,"vCpu":26,"systemRamGb":256,"region":"US West","affiliateUrl":"/go/lambdalabs","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0411,"fp8TflopsPerDollarHr":601.5,"fp16TflopsPerDollarHr":300.6,"monthlyRunRate730h":2401.7}},{"id":"modal-h100-80gb-serverless","provider":"Modal","providerType":"serverless","gpuModel":"NVIDIA H100 SXM5 (Serverless)","architecture":"Hopper","gpuCount":1,"vramGb":80,"totalVramGb":80,"interconnect":"NVLink 900GB/s","onDemandHourly":4.25,"spotHourly":null,"fp8Tflops":1979,"fp16Tflops":989,"vCpu":26,"systemRamGb":200,"region":"US Multi-Region","affiliateUrl":"/go/modal","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0531,"fp8TflopsPerDollarHr":465.6,"fp16TflopsPerDollarHr":232.7,"monthlyRunRate730h":3102.5}},{"id":"tensorwave-mi300x-single","provider":"TensorWave","providerType":"neocloud","gpuModel":"AMD Instinct MI300X","architecture":"CDNA 3","gpuCount":1,"vramGb":192,"totalVramGb":192,"interconnect":"Infinity Fabric 896GB/s","onDemandHourly":1.71,"spotHourly":1.25,"fp8Tflops":2610,"fp16Tflops":1300,"vCpu":32,"systemRamGb":256,"region":"US West (Las Vegas)","affiliateUrl":"/go/tensorwave","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0089,"fp8TflopsPerDollarHr":1526.3,"fp16TflopsPerDollarHr":760.2,"monthlyRunRate730h":1248.3},"spotMetrics":{"costPerGbVramHr":0.0065,"fp8TflopsPerDollarHr":2088,"fp16TflopsPerDollarHr":1040,"monthlyRunRate730h":912.5}},{"id":"runpod-mi300x-single","provider":"RunPod","providerType":"neocloud","gpuModel":"AMD Instinct MI300X","architecture":"CDNA 3","gpuCount":1,"vramGb":192,"totalVramGb":192,"interconnect":"Infinity Fabric 896GB/s","onDemandHourly":2.39,"spotHourly":1.75,"fp8Tflops":2610,"fp16Tflops":1300,"vCpu":32,"systemRamGb":256,"region":"US East","affiliateUrl":"/go/runpod","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0124,"fp8TflopsPerDollarHr":1092.1,"fp16TflopsPerDollarHr":543.9,"monthlyRunRate730h":1744.7},"spotMetrics":{"costPerGbVramHr":0.0091,"fp8TflopsPerDollarHr":1491.4,"fp16TflopsPerDollarHr":742.9,"monthlyRunRate730h":1277.5}},{"id":"crusoe-mi300x-single","provider":"Crusoe Cloud","providerType":"neocloud","gpuModel":"AMD Instinct MI300X","architecture":"CDNA 3","gpuCount":1,"vramGb":192,"totalVramGb":192,"interconnect":"Infinity Fabric 896GB/s","onDemandHourly":3.45,"spotHourly":null,"fp8Tflops":2610,"fp16Tflops":1300,"vCpu":32,"systemRamGb":256,"region":"us-east-1 (Zero Carbon)","affiliateUrl":"/go/crusoe","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.018,"fp8TflopsPerDollarHr":756.5,"fp16TflopsPerDollarHr":376.8,"monthlyRunRate730h":2518.5}},{"id":"coreweave-h200-141gb-sxm","provider":"CoreWeave","providerType":"neocloud","gpuModel":"NVIDIA H200 SXM","architecture":"Hopper","gpuCount":1,"vramGb":141,"totalVramGb":141,"interconnect":"NVLink 900GB/s","onDemandHourly":5.25,"spotHourly":3.85,"fp8Tflops":1979,"fp16Tflops":989,"vCpu":52,"systemRamGb":400,"region":"US East (Infiniband 3.2Tbps)","affiliateUrl":"/go/coreweave","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0372,"fp8TflopsPerDollarHr":377,"fp16TflopsPerDollarHr":188.4,"monthlyRunRate730h":3832.5},"spotMetrics":{"costPerGbVramHr":0.0273,"fp8TflopsPerDollarHr":514,"fp16TflopsPerDollarHr":256.9,"monthlyRunRate730h":2810.5}},{"id":"lambda-h200-141gb-sxm","provider":"Lambda Labs","providerType":"neocloud","gpuModel":"NVIDIA H200 SXM","architecture":"Hopper","gpuCount":1,"vramGb":141,"totalVramGb":141,"interconnect":"NVLink 900GB/s","onDemandHourly":4.1,"spotHourly":null,"fp8Tflops":1979,"fp16Tflops":989,"vCpu":52,"systemRamGb":400,"region":"US Central","affiliateUrl":"/go/lambdalabs","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0291,"fp8TflopsPerDollarHr":482.7,"fp16TflopsPerDollarHr":241.2,"monthlyRunRate730h":2993}},{"id":"hyperstack-b200-192gb","provider":"Hyperstack","providerType":"neocloud","gpuModel":"NVIDIA B200","architecture":"Blackwell","gpuCount":1,"vramGb":192,"totalVramGb":192,"interconnect":"NVLink 1.8TB/s","onDemandHourly":6.69,"spotHourly":4.9,"fp8Tflops":4500,"fp16Tflops":2250,"vCpu":64,"systemRamGb":512,"region":"EU Central","affiliateUrl":"/go/hyperstack","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0348,"fp8TflopsPerDollarHr":672.6,"fp16TflopsPerDollarHr":336.3,"monthlyRunRate730h":4883.7},"spotMetrics":{"costPerGbVramHr":0.0255,"fp8TflopsPerDollarHr":918.4,"fp16TflopsPerDollarHr":459.2,"monthlyRunRate730h":3577}},{"id":"neocloud-b300-288gb-spot","provider":"Specialized B300 Neocloud","providerType":"neocloud","gpuModel":"NVIDIA B300 (Ultra-VRAM)","architecture":"Blackwell","gpuCount":1,"vramGb":288,"totalVramGb":288,"interconnect":"NVLink 1.8TB/s","onDemandHourly":7.5,"spotHourly":0.5,"fp8Tflops":4500,"fp16Tflops":2250,"vCpu":96,"systemRamGb":1024,"region":"US West Tier-3","affiliateUrl":"/go/runpod","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.026,"fp8TflopsPerDollarHr":600,"fp16TflopsPerDollarHr":300,"monthlyRunRate730h":5475},"spotMetrics":{"costPerGbVramHr":0.0017,"fp8TflopsPerDollarHr":9000,"fp16TflopsPerDollarHr":4500,"monthlyRunRate730h":365}},{"id":"runpod-4x-rtx4090","provider":"RunPod","providerType":"neocloud","gpuModel":"4x NVIDIA GeForce RTX 4090","architecture":"Ada Lovelace","gpuCount":4,"vramGb":24,"totalVramGb":96,"interconnect":"PCIe Gen4","onDemandHourly":2.96,"spotHourly":1.76,"fp8Tflops":2640,"fp16Tflops":1320,"vCpu":64,"systemRamGb":256,"region":"US East","affiliateUrl":"/go/runpod","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0308,"fp8TflopsPerDollarHr":891.9,"fp16TflopsPerDollarHr":445.9,"monthlyRunRate730h":2160.8},"spotMetrics":{"costPerGbVramHr":0.0183,"fp8TflopsPerDollarHr":1500,"fp16TflopsPerDollarHr":750,"monthlyRunRate730h":1284.8}},{"id":"vast-8x-rtx4090","provider":"Vast.ai","providerType":"neocloud","gpuModel":"8x NVIDIA GeForce RTX 4090","architecture":"Ada Lovelace","gpuCount":8,"vramGb":24,"totalVramGb":192,"interconnect":"PCIe Gen4","onDemandHourly":4.72,"spotHourly":3.04,"fp8Tflops":5280,"fp16Tflops":2640,"vCpu":128,"systemRamGb":512,"region":"Global Mesh","affiliateUrl":"/go/vast-ai","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0246,"fp8TflopsPerDollarHr":1118.6,"fp16TflopsPerDollarHr":559.3,"monthlyRunRate730h":3445.6},"spotMetrics":{"costPerGbVramHr":0.0158,"fp8TflopsPerDollarHr":1736.8,"fp16TflopsPerDollarHr":868.4,"monthlyRunRate730h":2219.2}},{"id":"lambda-4x-a100-80gb","provider":"Lambda Labs","providerType":"neocloud","gpuModel":"4x NVIDIA A100 SXM4 (80GB)","architecture":"Ampere","gpuCount":4,"vramGb":80,"totalVramGb":320,"interconnect":"NVLink 600GB/s","onDemandHourly":9.96,"spotHourly":null,"fp8Tflops":0,"fp16Tflops":2496,"vCpu":128,"systemRamGb":512,"region":"US Central","affiliateUrl":"/go/lambdalabs","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0311,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":250.6,"monthlyRunRate730h":7270.8}},{"id":"azure-nd96isr-mi300x-8x","provider":"Microsoft Azure (ND MI300X v5)","providerType":"hyperscaler","gpuModel":"8x AMD Instinct MI300X","architecture":"CDNA 3","gpuCount":8,"vramGb":192,"totalVramGb":1536,"interconnect":"Infinity Fabric + 400Gbps IB","onDemandHourly":48,"spotHourly":28.8,"fp8Tflops":20880,"fp16Tflops":10400,"vCpu":96,"systemRamGb":1900,"region":"East US 2","affiliateUrl":"/go/azure","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0313,"fp8TflopsPerDollarHr":435,"fp16TflopsPerDollarHr":216.7,"monthlyRunRate730h":35040},"spotMetrics":{"costPerGbVramHr":0.0188,"fp8TflopsPerDollarHr":725,"fp16TflopsPerDollarHr":361.1,"monthlyRunRate730h":21024}},{"id":"aws-p5-48xl-h100-8x","provider":"AWS EC2 (p5.48xlarge)","providerType":"hyperscaler","gpuModel":"8x NVIDIA H100 SXM5","architecture":"Hopper","gpuCount":8,"vramGb":80,"totalVramGb":640,"interconnect":"NVLink + EFA 3200Gbps","onDemandHourly":98.32,"spotHourly":39.3,"fp8Tflops":15832,"fp16Tflops":7912,"vCpu":192,"systemRamGb":2048,"region":"us-east-1","affiliateUrl":"/go/aws","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.1536,"fp8TflopsPerDollarHr":161,"fp16TflopsPerDollarHr":80.5,"monthlyRunRate730h":71773.6},"spotMetrics":{"costPerGbVramHr":0.0614,"fp8TflopsPerDollarHr":402.8,"fp16TflopsPerDollarHr":201.3,"monthlyRunRate730h":28689}},{"id":"azure-nd96isr-h100-8x","provider":"Microsoft Azure (NDv5)","providerType":"hyperscaler","gpuModel":"8x NVIDIA H100 SXM5","architecture":"Hopper","gpuCount":8,"vramGb":80,"totalVramGb":640,"interconnect":"NVLink + Quantum-2 IB","onDemandHourly":105.8,"spotHourly":42.3,"fp8Tflops":15832,"fp16Tflops":7912,"vCpu":96,"systemRamGb":1900,"region":"East US 2","affiliateUrl":"/go/azure","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.1653,"fp8TflopsPerDollarHr":149.6,"fp16TflopsPerDollarHr":74.8,"monthlyRunRate730h":77234},"spotMetrics":{"costPerGbVramHr":0.0661,"fp8TflopsPerDollarHr":374.3,"fp16TflopsPerDollarHr":187,"monthlyRunRate730h":30879}},{"id":"gcp-a3-highgpu-8g","provider":"Google Cloud (A3 High)","providerType":"hyperscaler","gpuModel":"8x NVIDIA H100 SXM5","architecture":"Hopper","gpuCount":8,"vramGb":80,"totalVramGb":640,"interconnect":"NVLink + GPUDirect-TCPX","onDemandHourly":98.4,"spotHourly":38.8,"fp8Tflops":15832,"fp16Tflops":7912,"vCpu":208,"systemRamGb":1872,"region":"us-central1","affiliateUrl":"/go/gcp","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.1538,"fp8TflopsPerDollarHr":160.9,"fp16TflopsPerDollarHr":80.4,"monthlyRunRate730h":71832},"spotMetrics":{"costPerGbVramHr":0.0606,"fp8TflopsPerDollarHr":408,"fp16TflopsPerDollarHr":203.9,"monthlyRunRate730h":28324}},{"id":"coreweave-8x-h200","provider":"CoreWeave","providerType":"neocloud","gpuModel":"8x NVIDIA H200 SXM","architecture":"Hopper","gpuCount":8,"vramGb":141,"totalVramGb":1128,"interconnect":"NVLink + 3.2Tbps IB","onDemandHourly":42,"spotHourly":30.8,"fp8Tflops":15832,"fp16Tflops":7912,"vCpu":416,"systemRamGb":3200,"region":"US East","affiliateUrl":"/go/coreweave","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0372,"fp8TflopsPerDollarHr":377,"fp16TflopsPerDollarHr":188.4,"monthlyRunRate730h":30660},"spotMetrics":{"costPerGbVramHr":0.0273,"fp8TflopsPerDollarHr":514,"fp16TflopsPerDollarHr":256.9,"monthlyRunRate730h":22484}},{"id":"vast-rtx5090-single","provider":"Vast.ai","providerType":"neocloud","gpuModel":"NVIDIA GeForce RTX 5090","architecture":"Blackwell","gpuCount":1,"vramGb":32,"totalVramGb":32,"interconnect":"PCIe Gen5","onDemandHourly":0.67,"spotHourly":0.48,"fp8Tflops":1290,"fp16Tflops":260,"vCpu":16,"systemRamGb":64,"region":"US East","affiliateUrl":"/go/vast-ai","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0209,"fp8TflopsPerDollarHr":1925.4,"fp16TflopsPerDollarHr":388.1,"monthlyRunRate730h":489.1},"spotMetrics":{"costPerGbVramHr":0.015,"fp8TflopsPerDollarHr":2687.5,"fp16TflopsPerDollarHr":541.7,"monthlyRunRate730h":350.4}},{"id":"runpod-rtx5090-single","provider":"RunPod","providerType":"neocloud","gpuModel":"NVIDIA GeForce RTX 5090","architecture":"Blackwell","gpuCount":1,"vramGb":32,"totalVramGb":32,"interconnect":"PCIe Gen5","onDemandHourly":1.19,"spotHourly":0.69,"fp8Tflops":1290,"fp16Tflops":260,"vCpu":16,"systemRamGb":64,"region":"Global Community","affiliateUrl":"/go/runpod","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0372,"fp8TflopsPerDollarHr":1084,"fp16TflopsPerDollarHr":218.5,"monthlyRunRate730h":868.7},"spotMetrics":{"costPerGbVramHr":0.0216,"fp8TflopsPerDollarHr":1869.6,"fp16TflopsPerDollarHr":376.8,"monthlyRunRate730h":503.7}},{"id":"runpod-l40s-single","provider":"RunPod","providerType":"neocloud","gpuModel":"NVIDIA L40S","architecture":"Ada Lovelace","gpuCount":1,"vramGb":48,"totalVramGb":48,"interconnect":"PCIe Gen4","onDemandHourly":1.19,"spotHourly":0.79,"fp8Tflops":733,"fp16Tflops":366,"vCpu":16,"systemRamGb":64,"region":"US Secure Cloud","affiliateUrl":"/go/runpod","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0248,"fp8TflopsPerDollarHr":616,"fp16TflopsPerDollarHr":307.6,"monthlyRunRate730h":868.7},"spotMetrics":{"costPerGbVramHr":0.0165,"fp8TflopsPerDollarHr":927.8,"fp16TflopsPerDollarHr":463.3,"monthlyRunRate730h":576.7}},{"id":"runpod-a100-80gb-sxm","provider":"RunPod","providerType":"neocloud","gpuModel":"NVIDIA A100 SXM4 (80GB)","architecture":"Ampere","gpuCount":1,"vramGb":80,"totalVramGb":80,"interconnect":"NVLink (600 GB/s)","onDemandHourly":2.19,"spotHourly":1.49,"fp8Tflops":0,"fp16Tflops":312,"vCpu":16,"systemRamGb":128,"region":"US East Secure","affiliateUrl":"/go/runpod","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0274,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":142.5,"monthlyRunRate730h":1598.7},"spotMetrics":{"costPerGbVramHr":0.0186,"fp8TflopsPerDollarHr":0,"fp16TflopsPerDollarHr":209.4,"monthlyRunRate730h":1087.7}},{"id":"lambda-gh200-96gb","provider":"Lambda Labs","providerType":"neocloud","gpuModel":"NVIDIA GH200 Grace Hopper (96GB)","architecture":"Hopper + Grace","gpuCount":1,"vramGb":96,"totalVramGb":96,"interconnect":"NVLink-C2C (900 GB/s)","onDemandHourly":2.29,"spotHourly":null,"fp8Tflops":1979,"fp16Tflops":989,"vCpu":72,"systemRamGb":480,"region":"US Utah","affiliateUrl":"/go/lambdalabs","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0239,"fp8TflopsPerDollarHr":864.2,"fp16TflopsPerDollarHr":431.9,"monthlyRunRate730h":1671.7}},{"id":"vast-8x-rtx5090","provider":"Vast.ai","providerType":"neocloud","gpuModel":"8x NVIDIA GeForce RTX 5090","architecture":"Blackwell","gpuCount":8,"vramGb":32,"totalVramGb":256,"interconnect":"PCIe Gen5","onDemandHourly":5.84,"spotHourly":3.84,"fp8Tflops":10320,"fp16Tflops":2080,"vCpu":128,"systemRamGb":512,"region":"US Central","affiliateUrl":"/go/vast-ai","isDirectSponsor":false,"inStock":true,"metrics":{"costPerGbVramHr":0.0228,"fp8TflopsPerDollarHr":1767.1,"fp16TflopsPerDollarHr":356.2,"monthlyRunRate730h":4263.2},"spotMetrics":{"costPerGbVramHr":0.015,"fp8TflopsPerDollarHr":2687.5,"fp16TflopsPerDollarHr":541.7,"monthlyRunRate730h":2803.2}}]}