How is the monthly cost calculated?

Cost = (input tokens × input price + output tokens × output price) per million tokens, multiplied by your requests per day. Prices use the cheapest listed provider for each model.

Does this include cached-input discounts?

No — the calculator uses list prices. Providers like OpenAI and Anthropic offer 50–90% discounts on cached prompt tokens, so real costs for repeated prompts will be lower.

Which model is cheapest for production?

Sort the results table by monthly cost: open-weights-hosted models (DeepSeek, Llama, Qwen) and small frontier models typically land under $0.10 per 1M blended tokens.

LLM Token Cost Calculator

Enter your workload — average input/output tokens and requests per day — and instantly see what every model would cost per request, per day and per month.

Input tokens / requestOutput tokens / requestRequests / day

#	Model	Provider	$/Request	$/Day	$/Month	Score	$/1M in · out
Loading pricing data…

How the math works

Cost per request = (input_tokens ÷ 1M × input_price) + (output_tokens ÷ 1M × output_price). Monthly = daily × 30. Prices use the cheapest listed provider per model — cached-input discounts, batch APIs and commit tiers can lower real bills further. For per-task comparisons see Cost per Task.