AI budgets let you limit how much your AI agents can do in a calendar month. A budget caps the number of tool calls that AI agents are allowed to make, so you stay in control without watching every AI agent run.
What are AI budgets?
An AI budget is a monthly limit on the number of tool calls that AI agents may make. You can create as many budgets as you need and give each one a name and, optionally, a description. Using separate budgets, you can limit AI agents independently, for example per team or workflow.
A budget applies to the given calendar month and resets on the first day of every month. You do not need to reset it manually.
A budget limits how many actions AI agents take. It does not charge anything and it does not count tokens. LiveAgent does not charge for AI. Your AI agents run on your own account with an AI provider, and you pay the AI provider for what you use there.
What counts as a tool call?
Every action an AI agent performs on a ticket, for example reading the ticket messages, adding a tag, resolving the ticket or transferring it, is a tool call. Each tool call counts as one against the budget, regardless of which tool it is or how many tokens the AI uses. For example, an AI agent run that uses three tools to get ticket messages, get ticket notes and create a knowledge base article makes 3 tool calls.
To estimate how much of a budget a workflow will use, count how many actions the AI agent needs to take per ticket and multiply it by the number of tickets you expect it to process.
Default budget
A ready-to-use budget named Default with a limit of 300 000 tool calls per month is created for your account, so you can put AI agents to work immediately after setting up the AI agents feature. You can change its limit or create other budgets in the AI section of the configuration, under AI budgets.
Assigning a budget to an AI agent
Budgets are assigned to AI agents via rules that trigger them. When you add the "Run AI agent" action to a rule, you select both the AI agent and the budget its tool calls count against. Choosing a budget is required, so every AI agent run is tied to a specific budget. You also select a budget when you create an AI dispatch rule under Configuration > AI > AI Dispatch Rules.
What happens when a budget is used up
A run continues making tool calls until it finishes its work or until the budget reaches its monthly limit. When the limit is reached, further AI agent runs under that budget fail, and the run record shows the reason. The budget starts counting again on the first day of the next month. For more information on AI agent run outcomes, see Monitoring AI agent runs.
Tip: If AI agent runs frequently fail because a budget's limit is reached, either increase that budget's monthly limit or simplify your workflows so the AI needs fewer tool calls per ticket. Granting an AI agent only the tools it actually needs also helps.
Monitoring budget usage
The list of AI budgets shows how much of each budget has been used in the current month and how much remains. Usage is highlighted when a budget reaches 80% of its limit and again when it reaches the limit.
Monitoring AI agent runs
Every time an AI agent is triggered for a ticket, a run record that tracks its full lifecycle is created in the Reports > AI Agent Runs section. Each run record tracks the status, the number of tool calls it made and, if it failed, the reason. For a more detailed explanation, see the AI Agent Runs report article.