Agentic coding tasks consume 1000x more tokens than basic code reasoning, with high run-to-run variability, large model efficiency gaps, and weak self-prediction of costs.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
How Do AI Agents Spend Your Money? Analyzing and Predicting Token Consumption in Agentic Coding Tasks
Agentic coding tasks consume 1000x more tokens than basic code reasoning, with high run-to-run variability, large model efficiency gaps, and weak self-prediction of costs.