LLM Opex Every way to run an LLM, every cloud

Models

DeepSeek V3.1: price on every cloud

Open weights · 671 billion parameters · Mixture of experts · 163,840-token context · licence mit

Per token, by cloud

US dollars per million tokens, each cloud's cheapest region.
CloudInput, per 1M tokensOutput, per 1M tokensCheapest region
AWS$0.29$0.84eu-north-1 · 13 regions
Google Cloud$0.60$1.70global
Azure——not sold per token
OCI——not sold per token
Core42——not sold per token
Alibaba Cloud——not sold per token

Run it yourself

It needs 855.7 GB of GPU memory at fp8, the precision it ships in.

CloudCheapest instance that fitsGPUsAn hourA month, around the clock
AWSp5en.48xlarge8 × H200 141 GB$63.30$46,206
AzureStandard_ND96isr_MI300X_v58 × MI300X 192 GB$68.64$50,107
Google Clouda3-ultragpu-8g8 × H200 141 GB$84.81$61,909

More from DeepSeek

Similar size, open weights

Inside the UAE: no cloud sells it in-country.

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