Tool

LLM cost estimator.

Monthly cost for a conversational workload — pick a model, set the traffic, see the spend. Useful for go / no-go conversations before you commit to a build.

Model

$3/M input · $15/M output

Conversations per day

convos

Input tokens per conversation

tokens

Output tokens per conversation

tokens

Cached input share

Anthropic prompt caching · cached reads charged at ~10% of normal input rate

Need to size a prompt first? Use the token counter →