What’s really going on with AI token price deflation?
Token prices are collapsing. Here's why it matters for your AI spend forecast.

Why it matters
LLM inference pricing is deflating faster than model capability is improving, reshaping cloud economics and enterprise AI unit economics. This shifts the ROI calculus for deployments and changes how buyers should budget inference costs in 2027.
The key facts
5 to knowFT reporting on AI token price deflation as a substantive market phenomenon
Pricing pressure driven by increased competition, model efficiency gains, and surplus compute capacity
Enterprise inference costs declining while model capability remains competitive or improves
Implications for cloud economics and unit economics of AI workloads
Date: October 5, 2026
Go to the source
Financial Times Technologyft.com
Publisher excerpt: It’s a bit complicated and maybe not that surprising