OpenAI tests outcome-based pricing: pay only when AI succeeds
OpenAI is offering major enterprise customers an outcome-based pricing option, charging only when AI agents successfully complete tasks. Gary Marcus noted the move follows customer resistance to paying for a 62% failure rate.
2026-08-31 ~ 2026-09-01 · 4 related posts
- Scoop: OpenAI quietly testing outcome-based pricing for major customers — gaganghotra_ · 2026-08-31
- OpenAI tests 'pay for success' model for major enterprise customers — AccBalanced · 2026-08-31
- OpenAI Tests 'Pay When It Works' Pricing as Agent Fails 62% of Tasks — GaryMarcus · 2026-09-01
- OpenAI pilots outcome-based pricing, shifting AI business models — 机器之心 · 2026-09-01