LLM cost calculator
Enter monthly input and output token volume, choose a cloud model, and compare the estimated API bill with a purchased or rented GPU setup.
How to use this calculator
- Tokens are the basic units LLMs use to read and generate text. Roughly, 1 token ≈ ¾ of an English word. Input tokens are sent to the model; output tokens are what the model generates.
- Estimate your monthly input and output token volume based on expected user requests, prompt length, and response length. A typical support chatbot might use 10M input and 5M output tokens per month.
- Choose a cloud model and a GPU setup, then compare the estimated monthly costs. The result includes a recommendation and break-even analysis between cloud API and self-hosted inference.
Token Usage
Cloud API
Local Setup
Monthly Cost
Cloud API recommendedCloud API
$8.80
per month
Self-hosted
$94.44
per month
Cloud API saves $85.64/month
Cost Breakdown
Input tokens (10M @ $0.4/1M)$4.00
Output tokens (3M @ $1.6/1M)$4.80
Cloud total$8.80
GPU amortization (NVIDIA RTX 4090 ÷ 24mo)$75.00
Electricity (450W × 12h/day)$19.44
Local total$94.44
Rent NVIDIA RTX 4090
via RunPod — from $0.74/hr