Output and reasoning: what the terms mean
Savings levers
Output and reasoning
Cap outputs and set reasoning effort per task, so you pay for the answer, not the padding. A savings lever in the group “Model providers”.
The fix is a code change your team or coding agent makes.
Audon sizes it on your own cost data with one stated assumption and writes it as a potential saving. Levers that touch the same spend are sized in order, so the amounts add up.
A change you mark as done is measured from billing; an approved experiment makes it a verified saving.
Used in Agent Spend Optimization · How levers are sized
Related terms.
Service tier
Pay the fast or priority surcharge only where someone waits; run work nobody waits on at flex or batch rates.
Regional processing
Drop the regional or US-only processing surcharge where no contract requires it.
Model choice
Move to a newer, cheaper version of the same model, or route simpler requests to a cheaper model.
Tests and evals
Run tests, evals and CI on a cheaper model, a smaller sample or a nightly schedule.
Context size
Stop resending growing history and stay under long-context price thresholds.
Embeddings
Embed each document once (skip unchanged content) and run bulk embedding at batch rates.
Last updated