# WhichVM

> Compare every cloud VM in one place: specs, on-demand, spot and reserved prices in every region across AWS, Azure, Google Cloud, Akamai and Vultr. Find the right instance and the cheapest region to run it.

WhichVM is an independent, vendor-neutral comparison of 3,600+ virtual machine types across Amazon Web Services, Microsoft Azure, Google Cloud, Akamai Cloud (Linode), Vultr. Every instance has a page with its vCPUs, memory, processor, accelerators, storage and network, its hourly list price in every region (on-demand, spot, and 1- and 3-year commitments where offered), the cheapest region, and the closest equivalents on the other clouds. Prices are USD list prices before tax and discounts, refreshed daily from each provider's public price lists (last update 2026-11-12). Nobody pays for placement.

## Key pages

- [Amazon EC2](https://whichvm.com/aws): every instance type with specs and prices, filterable by region and pricing model
- [Azure Virtual Machines](https://whichvm.com/azure): every VM size with specs and prices, filterable by region and pricing model
- [Compute Engine](https://whichvm.com/gcp): every machine type with specs and prices, filterable by region and pricing model
- [Akamai Cloud Compute](https://whichvm.com/akamai): every plan with specs and prices, filterable by region and pricing model
- [Vultr Cloud Compute](https://whichvm.com/vultr): every plan with specs and prices, filterable by region and pricing model
- [Find a VM](https://whichvm.com/find): cheapest instances across all clouds for a given vCPU, memory, GPU, architecture and region
- [Compare](https://whichvm.com/compare): side-by-side comparison of up to four instances from any provider
- [Cloud GPU pricing](https://whichvm.com/gpu): every GPU instance by GPU model, with price per GPU-hour
- [Regions](https://whichvm.com/regions): every region of every provider, by continent
- [Methodology](https://whichvm.com/methodology): data sources, refresh schedule and how equivalents are matched

## Popular instances

- [t3.micro](https://whichvm.com/aws/t3.micro) (AWS): General purpose · 2 vCPU · 1 GiB · Intel Skylake E5 2686 v5 · Up to 5 Gigabit · $0.0104/hr on-demand in us-east-1 · spot $0.0043/hr · 1-yr $0.0065/hr
- [t3.medium](https://whichvm.com/aws/t3.medium) (AWS): General purpose · 2 vCPU · 4 GiB · Intel Skylake E5 2686 v5 · Up to 5 Gigabit · $0.0416/hr on-demand in us-east-1 · spot $0.0179/hr · 1-yr $0.0261/hr
- [t3.large](https://whichvm.com/aws/t3.large) (AWS): General purpose · 2 vCPU · 8 GiB · Intel Skylake E5 2686 v5 · Up to 5 Gigabit · $0.0832/hr on-demand in us-east-1 · spot $0.0409/hr · 1-yr $0.0522/hr
- [t4g.small](https://whichvm.com/aws/t4g.small) (AWS): General purpose · 2 vCPU · 2 GiB · AWS Graviton2 Processor · Up to 5 Gigabit · $0.0168/hr on-demand in us-east-1 · from $0.0112/hr in ap-south-1 · spot $0.0077/hr · 1-yr $0.0105/hr
- [m5.large](https://whichvm.com/aws/m5.large) (AWS): General purpose · 2 vCPU · 8 GiB · Intel Xeon Platinum 8175 · Up to 10 Gigabit · $0.096/hr on-demand in us-east-1 · spot $0.0434/hr · 1-yr $0.06/hr
- [m6i.large](https://whichvm.com/aws/m6i.large) (AWS): General purpose · 2 vCPU · 8 GiB · Intel Xeon 8375C (Ice Lake) · Up to 12.5 Gigabit · $0.096/hr on-demand in us-east-1 · spot $0.0425/hr · 1-yr $0.0635/hr
- [m7i.large](https://whichvm.com/aws/m7i.large) (AWS): General purpose · 2 vCPU · 8 GiB · Intel Xeon Scalable (Sapphire Rapids) · Up to 12.5 Gigabit · $0.1008/hr on-demand in us-east-1 · spot $0.0449/hr · 1-yr $0.0667/hr
- [m7g.large](https://whichvm.com/aws/m7g.large) (AWS): General purpose · 2 vCPU · 8 GiB · AWS Graviton3 Processor · Up to 12.5 Gigabit · $0.0816/hr on-demand in us-east-1 · from $0.0583/hr in ap-south-1 · spot $0.0367/hr · 1-yr $0.054/hr
- [m8g.large](https://whichvm.com/aws/m8g.large) (AWS): General purpose · 2 vCPU · 8 GiB · AWS Graviton4 Processor · Up to 12.5 Gigabit · $0.0898/hr on-demand in us-east-1 · from $0.0642/hr in ap-south-1 · 1-yr $0.0594/hr
- [c6i.large](https://whichvm.com/aws/c6i.large) (AWS): Compute optimized · 2 vCPU · 4 GiB · Intel Xeon 8375C (Ice Lake) · Up to 12.5 Gigabit · $0.085/hr on-demand in us-east-1 · from $0.0833/hr in ap-southeast-5 · spot $0.0392/hr · 1-yr $0.0562/hr
- [c7i.xlarge](https://whichvm.com/aws/c7i.xlarge) (AWS): Compute optimized · 4 vCPU · 8 GiB · Intel Xeon Scalable (Sapphire Rapids) · Up to 12.5 Gigabit · $0.1785/hr on-demand in us-east-1 · from $0.1749/hr in ap-southeast-5 · spot $0.0839/hr · 1-yr $0.1181/hr
- [c7g.large](https://whichvm.com/aws/c7g.large) (AWS): Compute optimized · 2 vCPU · 4 GiB · AWS Graviton3 Processor · Up to 12.5 Gigabit · $0.0725/hr on-demand in us-east-1 · from $0.0491/hr in ap-south-1 · spot $0.0341/hr · 1-yr $0.0478/hr
- [r6i.large](https://whichvm.com/aws/r6i.large) (AWS): Memory optimized · 2 vCPU · 16 GiB · Intel Xeon 8375C (Ice Lake) · Up to 12.5 Gigabit · $0.126/hr on-demand in us-east-1 · spot $0.0557/hr · 1-yr $0.0834/hr
- [r7i.large](https://whichvm.com/aws/r7i.large) (AWS): Memory optimized · 2 vCPU · 16 GiB · Intel Xeon Scalable (Sapphire Rapids) · Up to 12.5 Gigabit · $0.1323/hr on-demand in us-east-1 · spot $0.0591/hr · 1-yr $0.0875/hr
- [r7g.large](https://whichvm.com/aws/r7g.large) (AWS): Memory optimized · 2 vCPU · 16 GiB · AWS Graviton3 Processor · Up to 12.5 Gigabit · $0.1071/hr on-demand in us-east-1 · from $0.0751/hr in ap-south-1 · spot $0.045/hr · 1-yr $0.0708/hr
- [m5.xlarge](https://whichvm.com/aws/m5.xlarge) (AWS): General purpose · 4 vCPU · 16 GiB · Intel Xeon Platinum 8175 · Up to 10 Gigabit · $0.192/hr on-demand in us-east-1 · spot $0.0817/hr · 1-yr $0.121/hr
- [t2.micro](https://whichvm.com/aws/t2.micro) (AWS): General purpose · 1 vCPU · 1 GiB · Intel Xeon Family · Low to Moderate · $0.0116/hr on-demand in us-east-1 · spot $0.0043/hr · 1-yr $0.0072/hr
- [t2.medium](https://whichvm.com/aws/t2.medium) (AWS): General purpose · 2 vCPU · 4 GiB · Intel Xeon Family · Low to Moderate · $0.0464/hr on-demand in us-east-1 · spot $0.0203/hr · 1-yr $0.0287/hr
- [g4dn.xlarge](https://whichvm.com/aws/g4dn.xlarge) (AWS): GPU & accelerated · 4 vCPU · 16 GiB · 1 × NVIDIA T4 · 125 GB local · Up to 25 Gigabit · $0.526/hr on-demand in us-east-1 · spot $0.2204/hr · 1-yr $0.331/hr
- [g5.xlarge](https://whichvm.com/aws/g5.xlarge) (AWS): GPU & accelerated · 4 vCPU · 16 GiB · 1 × NVIDIA A10G · 250 GB local · Up to 10 Gigabit · $1.006/hr on-demand in us-east-1 · spot $0.5006/hr · 1-yr $0.6338/hr
- [g6.xlarge](https://whichvm.com/aws/g6.xlarge) (AWS): GPU & accelerated · 4 vCPU · 16 GiB · 1 × NVIDIA L4 · 250 GB local · Up to 10 Gigabit · $0.8048/hr on-demand in us-east-1 · spot $0.105/hr · 1-yr $0.5239/hr
- [p4d.24xlarge](https://whichvm.com/aws/p4d.24xlarge) (AWS): GPU & accelerated · 96 vCPU · 1,152 GiB · 8 × NVIDIA A100 40GB · 8,000 GB local · 400 Gigabit · $21.96/hr on-demand in us-east-1 · from $21.96/hr in us-east-2 · spot $13.07/hr · 1-yr $13.92/hr
- [p5.48xlarge](https://whichvm.com/aws/p5.48xlarge) (AWS): GPU & accelerated · 192 vCPU · 2 TiB · 8 × NVIDIA H100 · 30,720 GB local · 3200 Gigabit · $55.04/hr on-demand in us-east-1 · spot $57.76/hr
- [B1s](https://whichvm.com/azure/b1s) (Azure): General purpose · 1 vCPU · 1 GiB · Intel Xeon · 4 GB local · $0.0104/hr on-demand in eastus
- [B2s](https://whichvm.com/azure/b2s) (Azure): General purpose · 2 vCPU · 4 GiB · Intel Xeon · 8 GB local · $0.0416/hr on-demand in eastus
- [B2ms](https://whichvm.com/azure/b2ms) (Azure): General purpose · 2 vCPU · 8 GiB · Intel Xeon · 16 GB local · $0.0832/hr on-demand in eastus
- [D2s v5](https://whichvm.com/azure/d2s-v5) (Azure): General purpose · 2 vCPU · 8 GiB · Intel Xeon Platinum 8370C (Ice Lake) · $0.096/hr on-demand in eastus · spot $0.0203/hr · 1-yr $0.0592/hr
- [D4s v5](https://whichvm.com/azure/d4s-v5) (Azure): General purpose · 4 vCPU · 16 GiB · Intel Xeon Platinum 8370C (Ice Lake) · $0.192/hr on-demand in eastus · spot $0.0405/hr · 1-yr $0.1185/hr
- [D2as v5](https://whichvm.com/azure/d2as-v5) (Azure): General purpose · 2 vCPU · 8 GiB · AMD EPYC 7763v (Milan) · $0.086/hr on-demand in eastus · from $0.0556/hr in centralindia · spot $0.0182/hr · 1-yr $0.0531/hr
- [D2s v6](https://whichvm.com/azure/d2s-v6) (Azure): General purpose · 2 vCPU · 8 GiB · Intel Xeon Platinum 8573C (Emerald Rapids) · $0.101/hr on-demand in eastus · spot $0.0213/hr · 1-yr $0.0624/hr
- [D2ps v6](https://whichvm.com/azure/d2ps-v6) (Azure): General purpose · 2 vCPU · 8 GiB · Azure Cobalt 100 · $0.0702/hr on-demand in eastus · from $0.0462/hr in centralindia · spot $0.0306/hr · 1-yr $0.0414/hr
- [E4s v5](https://whichvm.com/azure/e4s-v5) (Azure): Memory optimized · 4 vCPU · 32 GiB · Intel Xeon Platinum 8370C (Ice Lake) · $0.252/hr on-demand in eastus · spot $0.0531/hr · 1-yr $0.1486/hr
- [F4s v2](https://whichvm.com/azure/f4s-v2) (Azure): Compute optimized · 4 vCPU · 8 GiB · Intel Xeon Platinum 8168 / 8272CL · 32 GB local · $0.169/hr on-demand in eastus · spot $0.0379/hr
- [NC4as T4 v3](https://whichvm.com/azure/nc4as-t4-v3) (Azure): GPU & accelerated · 4 vCPU · 28 GiB · 1 × NVIDIA T4 · 180 GB local · $0.526/hr on-demand in eastus · spot $0.1641/hr · 1-yr $0.3092/hr
- [NC40ads H100 v5](https://whichvm.com/azure/nc40ads-h100-v5) (Azure): GPU & accelerated · 40 vCPU · 320 GiB · 1 × NVIDIA H100 NVL · 3,576 GB local · $6.98/hr on-demand in eastus · spot $1.418/hr · 1-yr $5.235/hr
- [ND96isr H100 v5](https://whichvm.com/azure/nd96isr-h100-v5) (Azure): GPU & accelerated · 96 vCPU · 1,900 GiB · 8 × NVIDIA H100 · 28,000 GB local · $98.32/hr on-demand in eastus · spot $18.17/hr · 1-yr $62.92/hr
- [e2-micro](https://whichvm.com/gcp/e2-micro) (Google Cloud): General purpose · 2 vCPU · 1 GiB · AMD Rome, Intel Broadwell, Intel Haswell, Intel Skylake · Up to 1 Gbps · $0.00838/hr on-demand in us-central1 · spot $0.00503/hr · 1-yr $0.00528/hr
- [e2-small](https://whichvm.com/gcp/e2-small) (Google Cloud): General purpose · 2 vCPU · 2 GiB · AMD Rome, Intel Broadwell, Intel Haswell, Intel Skylake · Up to 2 Gbps · $0.0168/hr on-demand in us-central1 · spot $0.0101/hr · 1-yr $0.0106/hr
- [e2-medium](https://whichvm.com/gcp/e2-medium) (Google Cloud): General purpose · 2 vCPU · 4 GiB · AMD Rome, Intel Broadwell, Intel Haswell, Intel Skylake · Up to 2 Gbps · $0.0335/hr on-demand in us-central1 · spot $0.0201/hr · 1-yr $0.0211/hr
- [e2-standard-2](https://whichvm.com/gcp/e2-standard-2) (Google Cloud): General purpose · 2 vCPU · 8 GiB · AMD Rome, Intel Broadwell, Intel Haswell, Intel Skylake · Up to 4 Gbps · $0.067/hr on-demand in us-central1 · spot $0.0402/hr · 1-yr $0.0422/hr
- [e2-standard-4](https://whichvm.com/gcp/e2-standard-4) (Google Cloud): General purpose · 4 vCPU · 16 GiB · AMD Rome, Intel Broadwell, Intel Haswell, Intel Skylake · Up to 8 Gbps · $0.134/hr on-demand in us-central1 · spot $0.0804/hr · 1-yr $0.0844/hr
- [n2-standard-2](https://whichvm.com/gcp/n2-standard-2) (Google Cloud): General purpose · 2 vCPU · 8 GiB · Intel Cascade Lake, Intel Ice Lake · Up to 10 Gbps · $0.0971/hr on-demand in us-central1 · spot $0.0583/hr · 1-yr $0.0612/hr
- [n2-standard-4](https://whichvm.com/gcp/n2-standard-4) (Google Cloud): General purpose · 4 vCPU · 16 GiB · Intel Cascade Lake, Intel Ice Lake · Up to 10 Gbps · $0.1942/hr on-demand in us-central1 · spot $0.1165/hr · 1-yr $0.1224/hr
- [n2d-standard-2](https://whichvm.com/gcp/n2d-standard-2) (Google Cloud): General purpose · 2 vCPU · 8 GiB · AMD Milan, AMD Rome · Up to 10 Gbps · $0.0845/hr on-demand in us-central1 · from $0.0558/hr in asia-south1 · spot $0.0428/hr · 1-yr $0.0532/hr
- [n4-standard-4](https://whichvm.com/gcp/n4-standard-4) (Google Cloud): General purpose · 4 vCPU · 16 GiB · Intel Emerald Rapids · Up to 10 Gbps · $0.1814/hr on-demand in us-central1 · spot $0.1019/hr · 1-yr $0.1143/hr
- [c3-standard-4](https://whichvm.com/gcp/c3-standard-4) (Google Cloud): General purpose · 4 vCPU · 16 GiB · Intel Sapphire Rapids · Up to 23 Gbps · $0.2016/hr on-demand in us-central1 · spot $0.0754/hr · 1-yr $0.127/hr
- [c4-standard-4](https://whichvm.com/gcp/c4-standard-4) (Google Cloud): General purpose · 4 vCPU · 15 GiB · Intel Emerald Rapids, Intel Granite Rapids · Up to 10 Gbps · $0.1977/hr on-demand in us-central1 · spot $0.1183/hr · 1-yr $0.1245/hr
- [c4a-standard-4](https://whichvm.com/gcp/c4a-standard-4) (Google Cloud): General purpose · 4 vCPU · 16 GiB · Google Axion · Up to 23 Gbps · $0.1796/hr on-demand in us-central1 · spot $0.0932/hr · 1-yr $0.1184/hr
- [t2d-standard-1](https://whichvm.com/gcp/t2d-standard-1) (Google Cloud): General purpose · 1 vCPU · 4 GiB · AMD Milan · Up to 10 Gbps · $0.0422/hr on-demand in us-central1 · from $0.0279/hr in asia-south1 · spot $0.0253/hr · 1-yr $0.0266/hr
- [g2-standard-4](https://whichvm.com/gcp/g2-standard-4) (Google Cloud): GPU & accelerated · 4 vCPU · 16 GiB · 1 × NVIDIA L4 · Up to 10 Gbps · $0.7068/hr on-demand in us-central1 · from $0.7045/hr in us-east4 · spot $0.423/hr · 1-yr $0.4453/hr
- [a2-highgpu-1g](https://whichvm.com/gcp/a2-highgpu-1g) (Google Cloud): GPU & accelerated · 12 vCPU · 85 GiB · 1 × NVIDIA A100 40GB · Up to 24 Gbps · $3.673/hr on-demand in us-central1 · spot $1.989/hr · 1-yr $2.314/hr
- [a3-highgpu-8g](https://whichvm.com/gcp/a3-highgpu-8g) (Google Cloud): GPU & accelerated · 208 vCPU · 1,872 GiB · 8 × NVIDIA H100 · 6,000 GB local · Up to 800 Gbps · $88.49/hr on-demand in us-central1 · spot $52.70/hr · 1-yr $61.38/hr
- [Nanode 1GB](https://whichvm.com/akamai/g6-nanode-1) (Akamai): General purpose · 1 vCPU · 1 GiB · AMD EPYC · 25 GB local · 1 Gbps outbound · $0.0075/hr on-demand in us-east
- [Linode 2GB](https://whichvm.com/akamai/g6-standard-1) (Akamai): General purpose · 1 vCPU · 2 GiB · AMD EPYC · 50 GB local · 2 Gbps outbound · $0.018/hr on-demand in us-east
- [Linode 4GB](https://whichvm.com/akamai/g6-standard-2) (Akamai): General purpose · 2 vCPU · 4 GiB · AMD EPYC · 80 GB local · 4 Gbps outbound · $0.036/hr on-demand in us-east
- [Dedicated 4GB](https://whichvm.com/akamai/g6-dedicated-2) (Akamai): General purpose · 2 vCPU · 4 GiB · AMD EPYC · 80 GB local · 4 Gbps outbound · $0.054/hr on-demand in us-east
- [Dedicated 8GB](https://whichvm.com/akamai/g6-dedicated-4) (Akamai): General purpose · 4 vCPU · 8 GiB · AMD EPYC · 160 GB local · 5 Gbps outbound · $0.108/hr on-demand in us-east
- [Linode 24GB](https://whichvm.com/akamai/g7-highmem-1) (Akamai): Memory optimized · 2 vCPU · 24 GiB · AMD EPYC · 20 GB local · 5 Gbps outbound · $0.09/hr on-demand in us-east
- [vc2-1c-1gb](https://whichvm.com/vultr/vc2-1c-1gb) (Vultr): General purpose · 1 vCPU · 1 GiB · Intel Xeon · 25 GB local · $0.007/hr on-demand in ewr
- [vc2-1c-2gb](https://whichvm.com/vultr/vc2-1c-2gb) (Vultr): General purpose · 1 vCPU · 2 GiB · Intel Xeon · 55 GB local · $0.014/hr on-demand in ewr
- [vc2-2c-4gb](https://whichvm.com/vultr/vc2-2c-4gb) (Vultr): General purpose · 2 vCPU · 4 GiB · Intel Xeon · 80 GB local · $0.027/hr on-demand in ewr
- [vhf-1c-2gb](https://whichvm.com/vultr/vhf-1c-2gb) (Vultr): General purpose · 1 vCPU · 2 GiB · Intel Xeon · 64 GB local · $0.016/hr on-demand in ewr
- [vhp-2c-4gb-amd](https://whichvm.com/vultr/vhp-2c-4gb-amd) (Vultr): General purpose · 2 vCPU · 4 GiB · AMD EPYC · 100 GB local · $0.033/hr on-demand in ewr
- [voc-g-2c-8gb-50s-amd](https://whichvm.com/vultr/voc-g-2c-8gb-50s-amd) (Vultr): General purpose · 2 vCPU · 8 GiB · AMD EPYC · 50 GB local · $0.082/hr on-demand in ewr

## GPU pricing

- [NVIDIA B200 cloud pricing](https://whichvm.com/gpu/b200): every instance with NVIDIA B200 GPUs across clouds, price per GPU-hour
- [NVIDIA H200 cloud pricing](https://whichvm.com/gpu/h200): every instance with NVIDIA H200 GPUs across clouds, price per GPU-hour
- [NVIDIA H100 cloud pricing](https://whichvm.com/gpu/h100): every instance with NVIDIA H100 GPUs across clouds, price per GPU-hour
- [NVIDIA A100 80GB cloud pricing](https://whichvm.com/gpu/a100-80gb): every instance with NVIDIA A100 80GB GPUs across clouds, price per GPU-hour
- [NVIDIA A100 40GB cloud pricing](https://whichvm.com/gpu/a100): every instance with NVIDIA A100 40GB GPUs across clouds, price per GPU-hour
- [NVIDIA L40S cloud pricing](https://whichvm.com/gpu/l40s): every instance with NVIDIA L40S GPUs across clouds, price per GPU-hour
- [NVIDIA L4 cloud pricing](https://whichvm.com/gpu/l4): every instance with NVIDIA L4 GPUs across clouds, price per GPU-hour
- [NVIDIA A10G cloud pricing](https://whichvm.com/gpu/a10g): every instance with NVIDIA A10G GPUs across clouds, price per GPU-hour
- [NVIDIA A10 cloud pricing](https://whichvm.com/gpu/a10): every instance with NVIDIA A10 GPUs across clouds, price per GPU-hour
- [NVIDIA T4 cloud pricing](https://whichvm.com/gpu/t4): every instance with NVIDIA T4 GPUs across clouds, price per GPU-hour
- [NVIDIA V100 cloud pricing](https://whichvm.com/gpu/v100): every instance with NVIDIA V100 GPUs across clouds, price per GPU-hour

## Machine-readable data

- [Search index (JSON)](https://whichvm.com/data/search.json): every instance across providers with vCPU, memory, accelerator and reference on-demand price
- Amazon EC2: https://whichvm.com/data/aws/instances.json (full specs) and https://whichvm.com/data/aws/prices/{region}.json (hourly USD per instance), e.g. https://whichvm.com/data/aws/prices/us-east-1.json
- Azure Virtual Machines: https://whichvm.com/data/azure/instances.json (full specs) and https://whichvm.com/data/azure/prices/{region}.json (hourly USD per instance), e.g. https://whichvm.com/data/azure/prices/eastus.json
- Compute Engine: https://whichvm.com/data/gcp/instances.json (full specs) and https://whichvm.com/data/gcp/prices/{region}.json (hourly USD per instance), e.g. https://whichvm.com/data/gcp/prices/us-central1.json
- Akamai Cloud Compute: https://whichvm.com/data/akamai/instances.json (full specs) and https://whichvm.com/data/akamai/prices/{region}.json (hourly USD per instance), e.g. https://whichvm.com/data/akamai/prices/us-east.json
- Vultr Cloud Compute: https://whichvm.com/data/vultr/instances.json (full specs) and https://whichvm.com/data/vultr/prices/{region}.json (hourly USD per instance), e.g. https://whichvm.com/data/vultr/prices/ewr.json

## Optional

- [All 3,603 instances, condensed](https://whichvm.com/llms-full.txt)