Cisco Introduces AI Tokenomics Capability to Track and Forecast AI Costs

Cisco has introduced a Tokenomics capability within Splunk Agent Observability to help organisations monitor AI token consumption, attribute expenditure and forecast costs as they expand their use of artificial intelligence.

AI Tokenomics focuses on tracking the resources consumed by AI applications and understanding how that expenditure relates to business outcomes. The new capability combines usage, cost and productivity data to support budgeting and optimisation decisions.

“Understanding how AI resources are being used and what they are costing is becoming increasingly important as organizations scale their AI investments,” said Ahmed El Saadi, Vice President MEA & TRC at Splunk, a Cisco Company. “Tokenomics gives teams the insights they need to better understand consumption, manage spend and connect AI investments to business outcomes.”

Tracking usage and forecasting expenditure
The Tokenomics capability provides organisations with several tools to monitor and manage AI consumption. It enables teams to track token expenditure in real time and examine coding agent usage, costs, adoption and productivity at both team and individual levels. This visibility is intended to help organisations understand how AI tools are being used and where spending is concentrated.

The capability also consolidates consumption data across multiple AI tools and providers. Organisations can view employee usage of tools such as Claude Code and Codex alongside other AI coding applications through a centralised view of expenditure.

For cost planning, the system uses the Cisco Deep Time Series Model to forecast AI spending before the end of a billing period. These projections can help organisations anticipate expenditure and adjust budgets or usage accordingly.

The resulting insights may also help identify opportunities to route workloads to more cost-effective AI models, depending on the requirements of each task. The Tokenomics capability is available through Cisco Cloud Control and Splunk Agent Observability, both on-premises and through Splunk Observability Cloud. It forms part of the platform’s wider functionality for evaluating AI agent behaviour, monitoring performance across the AI stack and managing token consumption and associated costs.