CloudTopHosts is operated by WiserWeb, which also sells WiserWebCloud (listed here). Tables are ordered by advertised price. Sponsored placements are labelled and never change that order. How this works
CloudTopHosts

Crusoe

Categories
GPU clouds, Inference hosts
Owned by
Crusoe Energy Systems (private)
HQ
US
Website
https://crusoe.ai/cloud/managed-inference
Pricing page
https://crusoe.ai/cloud/pricing
Offers
GPU, managed k8s, object storage
Fee model
per_token+per_second_gpu
Fees
Serverless per 1M tokens (input/output/cached); Self-Serve Deployments (dedicated) H100 $5.50/h, H200 $6.00/h; Provisioned Throughput via AI Model Units (contact sales)
Free tier
no
OpenAI-compatible API
yes
Bring your own key
no
Notes
No egress charges on Crusoe Cloud. Part of Crusoe Intelligence Foundry.
Sources
https://crusoe.ai/cloud/pricing collected 2026-09-18

LLM prices (12 models, USD per 1M tokens)

ModelInputOutputCached inputContextSource
DeepSeek V4 Flash 0423 $0.14$0.28$0.03 src
DeepSeek V4 Pro 0423 $1.74$3.48$0.15 src
Gemma 4 31B $0.14$0.40$0.14 src
GLM-5.3 $1.40$4.40$0.26 src
GLM-5.3 Flash $0.15$0.50$0.03 src
gpt-oss-120b $0.05$0.20$0.05 src
Kimi K2.6 $0.70$3.50$0.35 src
Nemotron 3 Nano 30B $0.05$0.20$0.03 src
Nemotron 3 Super 120B A12B $0.30$2.40$0.15 src
Nemotron-3-Nano-Omni-30B $0.30$1.83$0.30 src
NVIDIA Nemotron 3 Ultra (Preview) $1$3.20$0.25 src
NVIDIA Nemotron 3.5 Lightning 30B A3B $0.05$0.20$0.03 src

GPU rates

ProviderGPUGPUsVRAM/GPUBillingPer GPU-hrInstance /hrRegionSource
CrusoeL40S 148 GBon demand $1.50$1.50 src
CrusoeA100 80GB PCIe 180 GBon demand $2$2 src
CrusoeA100 80GB SXM 180 GBon demand $2.30$2.30 src
CrusoeMI300X 1192 GBon demand $3.45$3.45 src
CrusoeH100 SXM 180 GBon demand $3.90$3.90 src
CrusoeH200 1141 GBon demand $4.29$4.29 src