Token Burn
Token waste caused by a client sending the entire session history to the LLM on every call.
Token Burn is an architectural problem, not a user problem. Many AI applications are built with the simplest possible approach: append the user's message to the full conversation history and send everything to the model. This means that by turn 20, the model is re-reading turns 1-19 — including all the dead ends, corrections, and superseded decisions. By turn 50, the token cost per call has grown dramatically, and most of what the model is processing is historical noise. The user doesn't see this; they just see their bill climbing and the model's responses slowing down.
More in Reasoning Cost
Reasoning Tax
Extra model cost paid to compensate for context debt.
Token Bleed
Silent budget drain from re-onboarding — tokens spent re-explaining context the system should already hold.
Context Compensation
The model using additional reasoning to make sense of poor, bloated, or unstructured context.
Reasoning Inflation
More reasoning is required to extract the same signal from worse context.
Reasoning Load
The amount of interpretive work a model must perform before useful task execution begins.
Cognitive Drag
Friction introduced by poorly structured context.
Reasoning Efficiency
The model producing correct output with minimal interpretive overhead — the reward for clean context.
Token Leverage
Getting more useful output per token spent. The inverse of Token Bleed.