Plans and pricing
Thousands of teams use CloudPrice every month to cut their cloud bills, finding the right VM size, region, and pricing model across Azure, AWS, and GCP, plus pricing for 1900+ AI models from 100+ providers. Subscribe to unlock all features, while helping us build what's next.
Free
$0/month
- AWS, Azure and GCP
- Specs for each Instance
- Similar Instances
- Price/Performance comparison
- AI pricing (1900+ LLMs)
- Free AI API & token calculator
Advanced
$9.95/month
- View ALL best price regions
- Comparison for Savings Options
- Spot Price History > 3 months
- Programmatic API access
- Compare more than 2 Instances
- VMs recommendations
- GPU → AI model fit recommendations
- Price/Perf for Savings Options
Batch export
$100/month
Azure only
- Daily snapshots in CSV (gzip)
- Instance specs and pricing
- Full dataset, no throttling
- All Azure services
- All regions & currencies
Need the full pricing history? Data goes back to 2022 and is exported separately, on request: contact [email protected].
People from these companies are using CloudPrice.net
Features available with subscription
AI Model Pricing & Benchmarks Without Throttling
Compare pricing and benchmarks for 1900+ AI models from over 100 providers, completely free. Estimate costs with the token calculator or pull data from our public AI Pricing API (no key, CORS-enabled).
Need higher volumes or the full dataset without rate throttling? Contact us.
$ curl https://ai.cloudprice.net/api/v1/models/openai-gpt-5-5/pricing/calculate?input_tokens=1000000&output_tokens=500000
{
"data": {
"model_id": "openai-gpt-5-5",
"currency": "USD",
"result": {
"total_cost": 6.25,
"providers": [{ "provider_id": "openai", "region": "global" }]
}
}
}GPU → AI Model Fit Recommendations
For any Azure, AWS, or GCP GPU instance we rank the open-weight models that fit in its VRAM by Intelligence Index. CloudPrice is the only tool that links GPU instances to a catalog of 1900+ LLMs, so you size the box to the model instead of guessing.
| Model | Intelligence | ~VRAM |
|---|---|---|
 Qwen3 32B | 67.4 | 64 GB |
 Gemma 3 27B | 55.1 | 54 GB |
 Mistral Small 3 24B | 53.0 | 48 GB |
 gpt-oss 20B | 49.8 | 16 GB |
Spot Price History
Spot prices change constantly. With a subscription you can analyze the complete price history, not just the last 3 months, reaching back to 2022 for Azure and early 2024 for AWS to find the cheapest, most stable times to run.
Savings Options Comparison
Compare every savings option for an instance in one place and pick the cheapest commitment for your workload. On Azure that means On-Demand, Reservations, Savings Plans, and Spot; on GCP, sustained-use discounts and 1-3 year commitments; on AWS, all 26 pricing options. Since reserved pricing varies by region and is booked in advance, choosing the right region and size up front avoids costly changes later.
Programmatic API access
Everything on CloudPrice.net, and more, is available through our REST API. Grab an API key from the CloudPrice Developer Portal and pull pricing straight into your apps and scripts. Need the full dataset in bulk without rate limits? Our Batch Export Data option has you covered.
View ALL Best Price Regions
The free tier shows the four cheapest regions for each instance. A subscription opens up a chart of every region for any instance size, sorted by price, so you can pick the cheapest option closest to you.
The gap is real: the same Standard_D2_v4 runs 62% more in West Europe than in Norway West, and users save *17-20% on region choice alone.
* based on users' feedbacks
Advanced Comparison Tool
Put up to 5 instances side by side and compare specs line by line, or their pricing across regions, on one screen. You can do the same with AI models, comparing pricing and benchmarks across providers. Every comparison lives in the URL, so you can copy it and share with your team.
Price/Performance for Reserved 1y-3y
The public view shows price/performance across all regions, but only for pay-as-you-go. A subscription adds the same analysis for Reserved 1-3y and Savings Plans. Since pricing shifts by region and payment model, so does the ratio, helping you find the best-value VM for your reservations. For example, 24 × Standard_F2s_v2 is 9.5% more efficient than 1 × Standard_F48s_v2 at the same CPU and RAM.
Frequently asked questions
Plans and billing
How do I cancel my subscription?
You can cancel anytime from your billing settings — open it and click Change or cancel subscription, no need to contact us. After cancelling you keep full access to Advanced features until the end of your current billing period, and you won't be charged again.
How long is the free trial?
Every Advanced plan starts with a 14-day free trial. You won't be charged until the trial ends, and you can cancel any time during those 14 days from your billing settings to avoid being billed.
What's the difference between monthly and yearly?
Both plans unlock the exact same Advanced features. The yearly plan is billed once a year and saves you 33% compared to paying month to month. You can start on either and switch later from your account settings.
Do you offer refunds?
Because every plan begins with a 14-day free trial, you can fully evaluate CloudPrice before paying anything. If you run into a billing issue, just get in touch and we'll make it right.
Is the AI model pricing data really free?
Yes. Pricing and benchmarks for 1900+ AI models from 100+ providers, the token cost calculator, and our public AI Pricing API (no key, CORS-enabled) are all free. A subscription is only needed for the cloud optimization tools.
I have another question — how do I reach you?
We're happy to help. Email our team and we'll get back to you as soon as we can.
Batch export
How do I get the data after subscribing?
Your existing CloudPrice API key unlocks it: when the subscription activates we re-scope that key to cover the exports, so there is no separate onboarding step. Each export is one gzipped file behind an endpoint on the developer portal, called with the subscription key from your API keys page.
What exactly is in an export?
Two files per service: specifications (instance type, vCPU, memory, storage, GPU) and pricing, both CSV inside a gzip archive. Azure VMs carry On-Demand, Low Priority and Spot rates in every currency we publish; AWS covers EC2 and EC2 Spot, RDS, Redshift, OpenSearch, DocumentDB, ElastiCache and SageMaker; GCP covers Compute Engine.
Can I look at the data before I buy?
Yes. Every service on the batch export page has a downloadable sample next to its specification, no account needed. The paid feed has the same shape, refreshed daily.
Can I subscribe to just one cloud?
Yes. Each cloud is sold on its own at $100/month, and all three together are $300/month (or $1,800/year). Every batch subscription also carries Advanced features and 5 Advanced seats, so your team keeps using the site while your pipeline consumes the export.
How far back does the data go?
The subscription delivers current pricing from the day you start and refreshes it daily. The historical archive is a separate one-time export: Azure VMs from mid-2022, AWS from January 2024, GCP Compute Engine from March 2025. Tell us the range you need and we'll quote it: email support.
Can I upgrade from one cloud to all three?
Yes. Subscribe to the all-cloud export, then cancel the single-cloud one from your billing settings; the new subscription covers everything the old one did. If the billing periods overlap awkwardly, get in touch and we'll sort it out.
Are there rate limits on the export?
An export is a file, not a paged API: one request returns the whole dataset for a service, so you never walk through thousands of calls to assemble it. The per-minute limits that shape our interactive pricing API don't come into play here.