AI Token Cost Estimator
Estimate your monthly AI API cost from token usage with editable per-model pricing, a prompt-caching toggle and a cheapest-model comparison table.
Requests per month (auto): 3,000 = requests per day × 30 days.
Model pricing (per 1M tokens)
These are example prices only — edit to match your provider. All prices in USD.
| Model class | Input $ / 1M | Output $ / 1M |
|---|
Your cost for GPT-4o class
$0.00
Cost per request
$0.00
Cost per day
$0.00
Cost per month
0
Tokens per month
Cheapest-model comparison
Same usage, every model class — sorted by monthly cost.
| # | Model class | Monthly cost | Per request |
|---|
How to use
- Enter input tokens, output tokens and requests per day.
- Pick your primary model class and edit the pricing table to match your provider — the defaults are examples only.
- Toggle prompt caching on if your provider discounts cached input tokens by 50 percent.
- Compare all model classes in the sorted table and pick the cheapest fit for your workload.
- Use "Load demo data" to see a realistic example.