API planning
LLM API budget calculator
Start with a monthly budget and estimate the request capacity of one LLM workload. This reverses the usual cost calculation so you can test whether a product plan fits its spend limit.
Free · no sign-in · editable assumptions
Enter your budget and workload
Cost / request$0.0000
At these rates
Requests / month0
Before safety margin
Tokens / month0M
Input + output
Planning view: enter values to see request capacity.
Use a safety margin: the result spends the entire budget on the modeled request. In a real plan, reserve room for retries, peak traffic, longer responses, cached/batch pricing differences, and other infrastructure costs.
How the budget calculation works
First, the tool calculates the cost of one request from separately priced input and output tokens. It then divides the monthly budget by that request cost.
request capacity = monthly budget ÷ cost per request
For example, a $100 budget and a $0.0045 request estimate supports about 22,222 modeled requests before you add a safety margin. If the request mix changes, capacity changes too.
Make the estimate useful
- Use measured token usage from production logs when you have it.
- Run separate estimates for short, average, and long responses.
- Reserve budget for retries, moderation, tools, and traffic spikes.
- Compare models using the same workload assumptions and current published rates.
For a forward monthly-spend calculation, use the LLM API cost calculator. For a deeper explanation of pricing meters, read how LLM API pricing works.