INSIGHT AI

Source-linked text edition ·

Cost per task: Using cheap classifiers to route across Gemini models and thinking levels

The written, source-linked counterpart to this video.

Spoken scripts · Report and scripts (PDF)

Published edition1 story

Cost per task: Using cheap classifiers to route across Gemini models and thinking levels

The dominating metric is now cost per successful task: total spend across every loop iteration, tool call, retry and internal thinking pass divided by the number of tasks that pass verification.

To prevent this unnecessary overhead, consider adding a lightweight routing layer that can intercept routine decisions before they ever reach an expensive reasoning model.

Source evidence for Cost per task: Using cheap classifiers to route across Gemini models and thinking levels
Source-linked proof · original source

Video edition · Markdown edition

Source ledger