GPT 4oMini (on claygent) went from 0.000x/row to 0.00x/row, how come? This is a massive price hike and its not scaleable at all. The prompt is simple. Table
Credit costs can fluctuate based on the AI model selected and token usage. Older rows might have used a less expensive model, while newer rows employ more advanced models that cost more credits per operation.
For AI enrichments, Clay uses variable credit costs based on prompt complexity and token count. When a prompt requires more tokens, additional credits are consumed only when necessary, not for every row.
To manage costs:
** Check which AI model is selected before execution
** Use filters to target specific rows
* Review your credit usage regularly