Summary
Mayank Pant, billing solutions architect at Stripe, presented findings from two years of working with AI companies on pricing strategy at AI Engineer Europe 2026. The talk opens with a striking data point from Stripe’s own transaction data: the top 100 AI companies reached $20 million ARR in an average of 20 months — compared to 65 months for the top 100 SaaS companies, a roughly 3x acceleration that is creating pricing challenges as fast as it is creating revenue.
Stripe’s research surfaces three structural problems unique to AI pricing: a small share of power users (5–10%) can consume 80% of compute, external infrastructure costs fluctuate unpredictably, and 84% of AI business leaders say their pricing cannot keep pace with product releases. The industry response has been a rapid shift to hybrid pricing — a base subscription fee combined with usage-based scaling — which grew from just 6% adoption in 2024 to 41% today. Intercom, Lovable, ElevenLabs, OpenAI, and Anthropic all use hybrid pricing and all bill through Stripe.
Pant walks through a five-step framework for designing AI pricing: define value delivered, choose a charge metric aligned to that value (tokens vs. outcomes vs. credits), select a pricing model, build billing guardrails to protect customer trust, and instrument for rapid iteration. A particularly actionable insight is the use of credit bundles to abstract away API-call or token-level pricing, making costs legible to end users while preserving backend flexibility. The session is directly applicable to any team building a monetized AI product and struggling to keep pricing aligned with fast-moving feature development.
📺 Source: AI Engineer · Published May 01, 2026
🏷️ Format: Deep Dive
