LLM Opex Every way to run an LLM, every cloud

Clouds

AWS: LLM prices, GPUs and regions

AWS sells 17 of the compared models per token and prices 9 kinds of GPU for hosting your own, across 36 regions.

Models sold per token

ModelInput, per 1M tokensOutput, per 1M tokensCheapest region
Amazon Nova Lite$0.06$0.24me-central-1
Amazon Nova Pro$0.40$1.60me-central-1
Claude Haiku 4.5$1.10$5.50ap-northeast-1
DeepSeek V3.1$0.29$0.84eu-north-1
Llama 3.1 405B$2.40$2.40us-east-2
Llama 3.1 70B$0.72$0.72us-east-1
Llama 3.1 8B$0.22$0.22us-east-1
Llama 3.3 70B$0.72$0.72us-east-1
Llama 4 Maverick 17B$0.24$0.97us-east-1
Llama 4 Scout 17B$0.17$0.66us-east-1
Mistral 7B$0.15$0.20us-east-1
Mistral Large$0.25$0.75us-east-1
Mistral Small$1.00$3.00us-east-1
OpenAI gpt-oss 120B$0.075$0.30eu-north-1
OpenAI gpt-oss 20B$0.035$0.15eu-north-1
Qwen3 235B A22B$0.11$0.44eu-north-1
Qwen3 32B$0.075$0.30eu-north-1

GPUs for hosting your own

On demand.
GPUFrom, per GPU-hourCheapest instanceGPUs in itRegion
T4$0.526g4dn.xlarge1 × 16 GBus-east-1
L4$0.8048g6.xlarge1 × 24 GBus-east-1
A10G$1.01g5.xlarge1 × 24 GBus-east-1
L40S$1.86g6e.xlarge1 × 48 GBus-east-1
A100$2.74p4d.24xlarge8 × 40 GBus-east-1
H100$6.88p5.48xlarge8 × 80 GBus-east-1
H200$7.91p5en.48xlarge8 × 141 GBus-east-1
B200$14.24p6-b200.48xlarge8 × 179 GBus-east-1
B300$17.80p6-b300.48xlarge8 × 268 GBus-east-1

Regions with prices

af-south-1, ap-east-1, ap-east-2, ap-northeast-1, ap-northeast-2, ap-northeast-3, ap-south-1, ap-south-2, ap-southeast-1, ap-southeast-2, ap-southeast-3, ap-southeast-4, ap-southeast-5, ap-southeast-6, ap-southeast-7, ca-central-1, ca-west-1, eu-central-1, eu-central-2, eu-north-1, eu-south-1, eu-south-2, eu-west-1, eu-west-2, eu-west-3, il-central-1, me-central-1, me-south-1, mx-central-1, sa-east-1, us-east-1, us-east-2, us-gov-east-1, us-gov-west-1, us-west-1, us-west-2

List prices collected 5 October 2026, refreshed every three hours. No negotiated rates or taxes.