Lowering the price of frontier AI models causes consumption to increase far more than expected, demonstrating Jevons paradox.

AI inference efficiency will improve 5-10x every 12-18 months.

“They'll be five to ten times more efficient every 12 to 18 months.”

Grok 4.5 uses just one-third the tokens of GPT-5.5 or Fable while achieving a similar intelligence score.

“Grok 4.5 uses just one-third the amount of tokens as GPT-5.5 or Fable while achieving a similar score.”

Decagon's token usage per conversation has increased over time because more model calls are added to improve quality, contrary to the assumption that optimization reduces token usage.

“Actually over time, the number of tokens we're using per conversation has gone up because we're actually doing more model calls to make the quality better.”

RAMP observed a 21x increase in customer token spend, prompting the creation of new finance controls like token spend management.

“Token spend among RAMP customers has grown by 21 times.”

Token costs are doubling every 45 days.

“Right now, our token costs are doubling every 45.”

During the DeepSeek moment in early 2025, the same week Nebius stock dropped 40%, the company had its best sales week ever.

“Nebul stock went down 40% in one week or so in February I think it was February or March 2024 2025 and anecdotal story the same the same exact week we probably had the best week in sales.”

The per-token price of frontier models has not declined as much as the market expected, because frontier competition is effectively limited to GPT and Claude.

“Pure unit price is not going down as much as people expected it to go.”