Primitive 13 / Card

Quota purchase card

Segmented usage meter plus a three-tier upgrade strip — Starter, Studio, Scale — with current and recommended badges keyed by usage ratio. Three live load levels — low (30% used → stay), high (85% → Studio) and at-cap (100% → Scale).

Production answer

Quota purchase card is a reusable Oak Flats Muffler Men UI primitive with documented states, accessibility expectations, theme behavior, and implementation evidence.

Primary CTAReview Quota purchase card states
Generative search brief

Quota purchase card: Segmented usage meter plus a three-tier upgrade strip — Starter, Studio, Scale — with current and recommended badges keyed by usage ratio. Three live load levels — low (30% used → stay), high (85% → Studio) and at-cap (100% → Scale).

State 1 · low usage · stay on Starter
Upgrade tier

Replicate predictions

30% used
6 preds/minof 20 preds/min
Replicate predictions quota usage
  • Starter$0
    20 preds / min
    • Shared GPU pool
    • 20 predictions per minute
    • Community support
    CurrentRecommended
  • Studio$49
    60 preds / min
    • Dedicated GPU lane
    • 60 predictions per minute
    • Email + Slack support
  • Scale$199
    240 preds / min
    • Reserved A100 capacity
    • 240 predictions per minute
    • 24/7 oncall
State 2 · high usage · Studio recommended
Upgrade tier

Replicate predictions

85% used
17 preds/minof 20 preds/min
Replicate predictions quota usage
  • Starter$0
    20 preds / min
    • Shared GPU pool
    • 20 predictions per minute
    • Community support
    Current
  • Studio$49
    60 preds / min
    • Dedicated GPU lane
    • 60 predictions per minute
    • Email + Slack support
    Recommended
  • Scale$199
    240 preds / min
    • Reserved A100 capacity
    • 240 predictions per minute
    • 24/7 oncall
State 3 · at-cap · Scale recommended
Upgrade tier

Replicate predictions

100% used
20 preds/minof 20 preds/min
Replicate predictions quota usage
  • Starter$0
    20 preds / min
    • Shared GPU pool
    • 20 predictions per minute
    • Community support
    Current
  • Studio$49
    60 preds / min
    • Dedicated GPU lane
    • 60 predictions per minute
    • Email + Slack support
  • Scale$199
    240 preds / min
    • Reserved A100 capacity
    • 240 predictions per minute
    • 24/7 oncall
    Recommended