The input/output token costs for v0 errors are too high.
When I prompt for an error, the output tokens are near zero. This makes sense but the high input tokens are consuming way too many credits. After ten attempts, I’ll consume $10 with zero results.
Maybe the team can cap error tokens to $0.25 to minimize the imbalance.
