allocateAI agent cost calculator
Estimate what a multi-turn agent costs to run. Set the turns, prompt size, and volume; get cost per task and per month, with context growth and caching included.
Without prompt caching the same task would cost $0.0050. After caching, routing the routine turns to a cheaper model is the biggest saving left.
How it works
Common questions
Why does cost grow so fast with turns?
Every turn resends the system prompt, the tool definitions, and everything said so far. By turn six the model is rereading five turns of history, so input tokens grow with the square of the turn count. Doubling the turns roughly quadruples the input cost.
How much does prompt caching save?
A lot for agents: the system prompt and tool definitions are identical every turn, so they cache at roughly a tenth of the input price, and stable history prefixes partially cache too. This calculator applies both; without caching, multiply the input side by several times.
How do teams cut agent cost in practice?
Four levers, in order of impact: route routine turns to a cheap fast model and keep frontier models for the hard steps, keep system prompts short, summarize history past a few turns, and cache aggressively. Routing alone usually saves more than everything else combined.
What does a typical production task cost?
A six-turn support task on a fast model with caching lands between a tenth of a cent and a few cents. The same task on a top-tier reasoning model can cost fifty times more, which is why choosing a model per route matters.
Is this calculator free?
Yes, free and unlimited, in your browser, no account.
More free tools
Allocate runs agent fleets with per-agent metering, hard caps, and a model choice per route, so agent spend stays inside a number you set in advance.