TokenPad
Cost

Subagent Token Budget Planner

Split a token budget across subagents, with the overhead counted honestly.

One subagent per line: name | share of budget (%)

40 characters3 lines0 tokensor drop a file

Subagent BudgetExact
0Token change
Token change0no change
Input tokens0what you pasted
Output tokens0what you would send
Token cost of this result
Output tokens0
As input$0.00
× 100K requests$0.00

Everything on this page runs in your browser. Nothing you paste is transmitted, because there is no server here to transmit it to.

Result
 

Overhead scales with the number of agents

Every subagent carries its own system prompt and tool schemas. Six subagents at 1,200 tokens each is 7,200 tokens of budget spent before any of them reads anything.

The planner subtracts that first, because allocating percentages of a total that includes overhead produces budgets nobody can actually work within.

When to collapse the topology

If overhead exceeds roughly thirty percent of the budget, the design is paying more for coordination than for work. Fewer agents with broader scope is usually both cheaper and more reliable.

Multi-agent structures earn their cost when sub-tasks need genuinely different tools or different models. Splitting purely to give each agent a tidier prompt rarely pays.

Frequently asked questions

How do I decide the shares?
By how much context each subagent actually needs to read, not by how important it feels. A research step consuming documents needs most of the budget; a formatting step needs very little.
Does this account for the history each agent accumulates?
No — it allocates context capacity. For the compounding cost within a single agent’s loop, use the agent loop simulator alongside this.

More cost tools