When developers assign coding tasks to AI agents, the process involves multiple steps: planning, generating code, testing, and iterating to fix errors. A single task can consume 10 to 50 times more tokens than a simple chat prompt. As these agents become more integral to workflows, token usage—and costs—escalate.
This financial strain is already evident. Uber reportedly depleted its 2026 AI budget by April after expanding its use of Claude Code. GitHub Copilot shifted from a flat-rate model to metered credits. Gartner predicts that AI coding token costs could exceed average developer salaries by 2028.
