According to @hyperbolic_labs, Hyperbolic has expanded H200 availability at a $1.99 per hour on-demand rate with VM and RoCE networking, and instances can be spun up in minutes, source: Hyperbolic on X, Jan 16, 2026. According to @hyperbolic_labs, longer reservations with discounts and dedicated model hosting for guaranteed capacity are also available, source: Hyperbolic on X, Jan 16, 2026. According to @hyperbolic_labs, at $1.99 per hour the implied per-GPU cost is approximately $47.76 per day and about $1,433 per 30 days, while an 8-GPU allocation would be roughly $382 per day and $11,462 per 30 days, calculations based on the stated rate, source: Hyperbolic on X, Jan 16, 2026. According to @hyperbolic_labs, these concrete rates provide a public reference for AI compute pricing that traders can use to benchmark operating costs for AI infrastructure plays and crypto projects offering model inference or training, source: Hyperbolic on X, Jan 16, 2026.