HighSave ~€947/mo
Wrong model for the task
- Finding
- 13 people run routine work on the premium reasoning model.
- Evidence
- 65% of their spend is premium; overspend is measured against a cheaper model at list price.
- Suggested action
- Set the standard model as default for drafting and summaries.