GPU Cloud Cost Calculator
Compare GPU rental costs across AWS, GCP, Azure, Lambda Labs, and Vast.ai for ML training and inference workloads.
Did this tool work for you?
How to use this calculator
- 1
Select the GPU type that matches your workload requirements.
- 2
Enter the number of GPU-hours you expect to use per month.
- 3
Select your preferred cloud provider.
- 4
Compare the monthly cost and see the cheapest available provider for the same GPU.
Frequently asked questions
How is the cost computed?
Monthly cost = the hourly rate for the chosen GPU and provider × the hours you enter. Annual cost is monthly × 12. The tool also finds the lowest rate for the same GPU across the other listed providers and shows the savings versus that rate, (your rate − cheapest rate) × hours.
Where do the hourly rates come from?
From a fixed table in the code, commented as approximate on-demand pricing in US dollars per GPU-hour. There is no live lookup. Cloud GPU prices vary by region, instance size, commitment and availability and are revised often, so check each provider’s pricing page.
Are the hours GPU-hours or instance-hours?
GPU-hours. The rate is per single GPU, so a job on eight GPUs for ten hours is 80 hours. If you rent a multi-GPU instance, multiply instance hours by the number of GPUs.
Why does the RTX 3090 show "N/A" on AWS, GCP and Azure?
The table has no entry for that combination, so the tool returns a note suggesting Lambda Labs or Vast.ai instead of a price. It does not mean a provider can never offer the card; it means this tool has no rate stored.
Does it include spot, preemptible or reserved pricing?
No. There is one rate per GPU and provider and no discount modelling. If you plan to use discounted capacity, apply the discount to the hourly rate yourself by comparing with the provider’s published options.
What does "cheapest provider" compare?
Only the stored hourly rates for the same GPU among providers that have one. It ignores performance, memory configuration, region, support, storage, data transfer and reliability, so the cheapest listed rate is not necessarily the best fit.
Does the monthly cost assume the GPU runs all month?
No. You enter the hours. A GPU left on around the clock for a 30-day month is 720 hours, which is the figure used for the always-on examples on this page.
What is left out of the estimate?
Storage, snapshots, outbound data transfer, CPU and memory sizing, load balancers, taxes and support plans. Only the GPU hourly charge is modelled.
GPU cloud cost calculator: hourly rates, monthly bills and the cheapest listed provider
How the estimate works
Pick a GPU, a provider and the number of GPU-hours per month. The tool looks up an hourly rate in a built-in table, multiplies it by the hours for the monthly cost and by twelve for the annual cost, then scans the providers for the lowest rate on the same GPU. Savings versus cheapest equals (your rate − cheapest rate) × hours, so selecting the cheapest provider shows $0.00. Hours are GPU-hours: a job on eight GPUs for ten hours is 80. If a combination has no stored rate, the tool says so rather than guessing.
Rates stored in the code
These are the figures built into this tool, labelled in the code as approximate on-demand rates per GPU-hour in US dollars. Cloud GPU prices vary by region, instance size, commitment and availability, and are revised often, so treat the table as a rough starting point and check each provider’s pricing page.
| GPU | AWS | GCP | Azure | Lambda Labs | Vast.ai |
|---|---|---|---|---|---|
| A100 80GB | 4.10 | 4.50 | 4.20 | 2.49 | 2.00 |
| H100 80GB | 9.84 | 10.60 | 9.98 | 4.99 | 3.80 |
| A10G 24GB | 1.53 | 1.80 | 1.65 | 0.75 | 0.45 |
| V100 16GB | 3.06 | 2.55 | 3.20 | 1.50 | 0.70 |
| T4 16GB | 0.53 | 0.35 | 0.48 | 0.50 | 0.15 |
| RTX 3090 24GB | — | — | — | 0.99 | 0.35 |
Worked examples
GPU cost runs. The columns show Run, Hourly rate, Monthly, Annual and Cheapest listed, one line per case.
| Run | Hourly rate | Monthly | Annual | Cheapest listed |
|---|---|---|---|---|
| A100 80GB, AWS, 100 hours (defaults) | $4.10 | $410.00 | $4,920.00 | Vast.ai at $2.00/hr, $210.00/mo less |
| H100 80GB, AWS, 720 hours | $9.84 | $7,084.80 | $85,017.60 | Vast.ai at $3.80/hr, $4,348.80/mo less |
| T4, GCP, 720 hours | $0.35 | $252.00 | $3,024.00 | Vast.ai at $0.15/hr, $144.00/mo less |
| RTX 3090, AWS, 200 hours | N/A | N/A | N/A | Note suggests Lambda Labs or Vast.ai |
What 720 hours means
A GPU left running all month is 720 hours at 30 days, which is how the second and third runs were set. An always-on inference endpoint therefore pays for the whole month whether or not requests arrive, while a training job pays only for its run time. Sizing hours honestly, including idle time between experiments, usually moves the bill more than the choice of provider. If a run finishes early because you stop the instance, only the hours it was actually on belong in the field.
Costs outside the table
The rate covers the GPU only. Disk, snapshots, outbound data transfer, CPU and RAM sizing, load balancers, spot or reserved discounts, taxes and support plans are not included, and the table has no regions, so a quote for a specific region may differ. Where a provider has no stored listing for the card, as with the RTX 3090 on the three large clouds, the tool reports that instead of showing a price. Use the output to shortlist options, then price the exact configuration on the provider’s own calculator.
Learn more from an authoritative source:
OpenAI Platform DocsAI Chatbot ROI Calculator
Calculate the return on investment of deploying an AI chatbot for customer support, including payback period and annual ROI.
LLM Latency Calculator
Estimate expected LLM response latency — time-to-first-token and total response time — for planning real-time AI applications.
AI Token Counter
Estimate the number of tokens in your text for GPT-4, Claude, Gemini, and other LLMs. Useful for staying within context limits.
AI Prompt Cost Calculator
Calculate the cost of an AI API call based on input/output tokens and model pricing.
Words to Tokens Converter
Convert between words, characters, tokens, and pages for AI models and content planning.
AI API Budget Calculator
Plan your monthly AI API budget based on usage volume, model selection, and request patterns.
Results are estimates for informational purposes only and do not constitute professional financial, medical, legal, or technical advice. Read full disclaimer →