Goldman warns declining token prices may hinder investment growth
Hyperscaler return on invested capital hits an all-time low as cheap open-source models and plummeting token prices squeeze the economics of AI infrastructure spending
The price of running AI just fell off a cliff, and Goldman Sachs thinks that’s a problem for the companies spending hundreds of billions to build the infrastructure behind it.
Goldman’s Delta One trading desk issued a warning in September 2026 about the structural risks facing hyperscalers as the cost of AI model usage tokens collapses faster than demand can grow. The Silicon Data LLM Token Expenditure Index dropped 29% in August 2026, landing at a record low of $0.97 per million tokens. That’s more than 50% below its May 2026 peak of roughly $2.05.
The math isn’t mathing
Rich Privorotsky, head of Goldman’s Delta One desk, pointed to two converging forces crushing per-token pricing for cloud inference: the rapid improvement of proprietary models (which can do more with less) and the rising tide of open-source alternatives that undercut commercial offerings on price.
Goldman flagged a scenario where token prices fall 30% but usage only grows 10%, a gap that would eat directly into revenue from AI infrastructure investments. That’s not a hypothetical. It’s roughly what August’s numbers suggest is already happening.
The companies caught in this squeeze are the biggest names in tech: Meta, Microsoft, Amazon, and Alphabet. All four have committed to capital expenditure programs for AI infrastructure that Goldman described as unprecedented in the tech sector’s history.
The news moving money, markets, and the world—before your day starts.
Daily. Free. Join 34,000+ readers across crypto, finance, and policy.
Open source keeps winning on price
Models that once would have required expensive API calls to frontier labs can now be run locally or on cheaper cloud instances using open-weight alternatives. Every improvement in open-source capability puts downward pressure on what commercial providers can charge.
Meta’s Muse Spark 1.3 recently launched, while OpenAI’s Astra is on the horizon. Both are expected to intensify the pricing war.
Return on invested capital for hyperscalers has already dropped to an all-time low, according to Goldman’s analysis.
The token pricing collapse from $2.05 to $0.97 in roughly three months is the kind of move that forces strategic recalculation. Whether hyperscalers can engineer enough demand growth to outrun the price decline will likely define the next chapter of the AI investment cycle.