Kimi K3 Paused New Subscriptions in 48 Hours-Design Your AI Onboarding for That Day

Short summary

Kimi K3 paused new subscriptions within 48 hours of launch due to compute demand exceeding capacity. The author uses this as a case study for designing AI product onboarding that handles capacity limits gracefully, proposing a three-state flow (available, waitlist, existing-user priority) and a waitlist page contract. The piece also highlights capacity-based product splitting as a design strategy from day one.

• •AI onboarding needs a third state beyond available/full: a waitlist with clear expectations on why, when, and what happens next
• •Never let users discover capacity limits by failing—communicate before they start the signup flow
• •Consider splitting product features for capacity reasons, not just pricing or marketing, to enable independent scaling

Generated with AI, which can make mistakes.