TOTAL VOLUME:
$114.5b
24H VOL:
$142,891,824
24H TRANSACTIONS:
1,380,975,298
OPEN INTEREST:
$1,211,057,904
339,560
Markets across
34,155
events
MATCHED EVENTS:
4,672
PLATFORM COVERAGE:
5
Polymarket:
42%
VS.
Kalshi:
58%
Above $2.50
- Kalshi
Above $2.50 - Kalshi
1W
News
Positive
Negative
Neutral
Hover marker for details
Vol.
·
Resolves Sep 1, 2026
Time left: 04d:08h:16m
$
NVIDIA's H100 GPU is a high-performance accelerator used for AI and machine learning workloads. This event tracks whether the average hourly rental or compute price for H100 capacity reaches various price thresholds during August 2026, reflecting market demand and supply dynamics for premium GPU computing resources.
Resolution is determined by the arithmetic mean of hourly H100 compute prices reported by Ornn throughout August 2026, using USD values rounded to two decimal places. Each market outcome corresponds to a specific price threshold: the event resolves Yes if the monthly average exceeds the stated threshold ($2.00, $2.25, $2.50, $2.75, $3.00, $3.25, $3.50, $3.75, or $4.00 per hour). Data revisions made after August 2026 expires are not incorporated into the final resolution.
Prediction market odds on this market reflect crowdsourced expectations from traders with real money at stake, often diverging from published analyst forecasts. While traditional equity research and cloud pricing reports offer point estimates based on historical trends and vendor guidance, traders here incorporate broader signals—supply chain dynamics, competitive pricing pressure, and demand shifts—into live odds. Comparing the two reveals whether the market is pricing in more optimism or caution than consensus analyst views on H100 availability and cost trajectories through mid-2026.
On Kalshi, this market is priced through a continuous order-book mechanism where traders buy and sell shares representing different price ranges or outcomes for the H100's monthly average compute cost. On Kalshi, prices reflect that venue's order book, liquidity, and how traders price the outcome right now. The bid-ask spread reflects the current consensus and liquidity; tighter spreads indicate higher confidence and trading activity, while wider spreads suggest uncertainty or lower participation. As new information emerges about GPU supply, demand, or NVIDIA's pricing strategy, traders adjust their positions, moving the odds to reflect updated expectations.
This market resolves around Sep 1, 2026, once the August 2026 monthly average compute price for NVIDIA's H100 is verifiable from credible public sources. The outcome is determined by the actual pricing data available at that time, confirmed through industry reports, NVIDIA's official communications, or established cloud pricing indices. Traders' positions settle based on whether the realized price falls within the range they backed, with payouts distributed accordingly once the final figure is established and verified.
Several catalysts could shift odds significantly before resolution. NVIDIA's quarterly earnings calls and product roadmap announcements may signal supply or demand changes. Competitive GPU releases from AMD or other vendors could pressure H100 pricing downward. Cloud provider pricing announcements—from AWS, Google Cloud, or Azure—directly reflect H100 compute costs and would move the market sharply. Macroeconomic shifts affecting AI infrastructure spending, geopolitical supply chain disruptions, or changes in data center utilization rates could also reshape trader expectations around August 2026 pricing levels.