BIGMODEL?RUN THE MATHS
SUPPORT · EDITORIAL CLASSIFICATION · NOT A BENCHMARK

DO I NEED THE BIG MODEL FOR…

Check refund eligibility against fixed rules

Small-model jobNO — START SMALL

Rules belong in code; a small model can translate the conversation around them.

START WITH
THE WORK.

  1. 01

    The work is constrained enough to evaluate cheaply.

  2. 02

    A clear schema, rubric, or example carries much of the workload.

  3. 03

    Most failures can be detected before they become expensive.

RECOMMENDED SETUPSmall, fast model + schema + deterministic checks
NON-NEGOTIABLE

Make the policy engine authoritative and log overrides.

REASONING2/5
AMBIGUITY2/5
ERROR COST3/5
CONTEXT2/5
CALCULATE THIS WORKLOAD →OPEN THE INTERACTIVE CHECKER →