Version 2

H100 GPU spot prices remain roughly 40% below their May 2026 peak at ~$2.42/hr [1], while one-year H100 contract prices have climbed from a trough of ~$1.70/hr to ~$2.65/hr over the same period [3]. SemiAnalysis argues this divergence is a market maturation signal, not a demand-weakness signal [5]. The bearish read of spot price declines persists among broader market observers [2], though new voices are explicitly pushing back on the AI bubble framing [9].

Spot price indexes are the most visible public data on compute health, and their decline has sustained a bearish demand narrative. If contract prices — which reflect committed production workloads — are moving in the opposite direction, observers relying on spot data alone may be drawing inverted conclusions about where sustained AI investment is headed.

H100 GPU spot prices have declined to approximately $2.42 per hour as of late June 2026, sitting roughly 40% below their May peak [1]. This sustained drop has generated concern among AI observers and investors that appetite for compute is cooling — a reading amplified by multiple public spot price indexes and by market commentators framing the decline as evidence of stress in the AI compute boom [2].

SemiAnalysis, which publishes a proprietary neocloud survey, posted a five-tweet thread on June 26 arguing that spot price indexes are the wrong instrument for measuring AI demand health [1][3]. The firm draws a structural distinction between spot and on-demand markets — which serve proof-of-concept runs, one-off evaluations, burst workloads, and overflow capacity — and term contract markets, where buyers commit to planned, recurring, revenue-bearing production inference or training workloads [4]. One-year H100 contract prices on neoclouds have climbed from a trough of roughly $1.70/hr in late 2025 to approximately $2.65/hr as of June 2026 [3]. SemiAnalysis attributes this to serious buyers locking in term capacity for production workloads, pushing contract prices higher even as the spot market remains loose [5].

The firm's conclusion inverts the prevailing read: falling spot prices alongside rising contract prices signal market maturation — buyers shifting from exploratory, flexible capacity toward committed production deployment — not a cooling of AI ambition [5][6]. SemiAnalysis's LinkedIn GPU price increase alert [7] and a Perplexity summary reporting H100 rental prices surging nearly 40% [8] have extended that message beyond the original tweet thread. A separate LinkedIn post by Cristian Sion makes the anti-bubble case directly, asking why GPU prices for older hardware are rising if AI is in a bubble [9].

A complicating data point is the SiliconData H100 Hyperscaler Index for April 2026, which showed pricing in flat mode [10] — a different signal from both the spot price decline and the neocloud contract price rise. Whether hyperscaler contract pricing has since diverged from the neocloud trajectory remains unresolved.