AI resource consumption billing scenarios

Genesys Cloud implements a tiered AI resource consumption billing model that charges customers based on the highest-tier resource utilized during each interaction. The platform establishes distinct pricing structures across different AI resource types: bot flows are charged at 17 minutes per token for voice interactions or 51 sessions per token for digital channels, virtual agents are billed at 0.5 tokens per interaction, and agentic virtual agents are charged at 1.2 tokens per interaction. When multiple AI resources are deployed within a single interaction, Genesys Cloud applies the highest applicable resource rate to the entire engagement rather than charging separately for each component. For instance, interactions combining bot flows with virtual agents are charged at 0.5 tokens per interaction, while those incorporating bot flows with agentic virtual agents incur 1.2 tokens per interaction. Virtual agent interactions that include guides or Language Agent Model (LAM) capabilities are consistently charged at one token per interaction regardless of active utilization of these features. Billing calculations are based on interaction IDs for single-resource usage and apply the maximum resource tier rate for mixed-resource scenarios. Notably, Genesys charges the 1.2 token price for agentic virtual agents as soon as a flow visits a block containing this resource, even if the block is not executed during the flow. Bot flows cannot include multiple branches within a single flow, and the presence of an agentic virtual agent module prevents the reduced 0.5-token rate from applying. To optimize costs and manage AI resource consumption more effectively while maintaining transparency in billing calculations, Genesys recommends that customers call different bot flows from inbound call flows rather than consolidating multiple branches within a single flow structure.