Simple limits, honest metering.
A Standard run costs 1 unit, a Deep Agent optimization costs 5 — you see the cost before every run, and a run that fails, errors, or is refused never counts against you.
Plus
$3.99/month
For regular prompt work.
- 10 units a day
- 200 units a month
- Deep Agent optimization included
- Up to 2 Deep runs a day
Pro
$19.99/month
For prompt-heavy workflows.
- 50 units a day
- 1,000 units a month
- Deep Agent optimization included
- Up to 10 Deep runs a day
Prices are monthly; there is no annual plan yet. Unused units do not roll over.
Questions
- What counts as a unit?
- A Standard run consumes 1 unit; a Deep run consumes 5. The cost is shown before you start every run.
- What's the difference between Standard and Deep?
- Standard is a fast optimization with reasoning and no independent score. Deep starts from an optimized prompt, lets an Agent diagnose and improve it, and uses independent absolute scoring to drive bounded iterations and an explicit threshold decision.
- What happens when I hit a limit?
- Runs pause until the limit resets. Daily limits reset at 00:00 UTC; monthly limits reset with your billing cycle. Both limits apply — whichever runs out first.
- Do failed runs count?
- No. A run that errors, is cancelled, or is refused never counts against your limits.
- Who can see my results?
- Every result starts private. You can explicitly create an unlisted public link or publish to Explore, and you can revoke either at any time.