Hyperbolic raises GPU rates again — H100 SXM to $3.19/hr, floor claim now matches it
Hyperbolic raised H100 SXM to $3.19/GPU/hr (+10%) and H200 to $3.99 (+14%); its 'starting at' price now equals the H100 rate, replacing the old $0.20/GPU/hr floor.
H100 SXM $2.89/GPU/hr, H200 $3.49, B200 $5.99; marketplace advertised 'starting at $0.20/GPU/hr' with no matching GPU on the page.
H100 SXM $3.19/GPU/hr, H200 $3.99, B200 $5.99; marketplace now advertises 'starting at $3.19/GPU/hr' — the same figure as the H100 SXM rate.
Proof of change
Hyperbolic's pricing pages, as we captured them on two dates.
Showing the whole page as captured — scroll either panel, or open it at full size.
Show 4 other pages we compared
Showing the whole page as captured — scroll either panel, or open it at full size.
Showing the whole page as captured — scroll either panel, or open it at full size.
Showing the whole page as captured — scroll either panel, or open it at full size.
Showing the whole page as captured — scroll either panel, or open it at full size.
Both images are our own captures, taken on the dates shown. The pixels are unmodified — nothing is retouched, and where a region is highlighted the marker is drawn over the image, not into it. Open either capture to see it at full resolution.
Hyperbolic’s on-demand marketplace refreshed its published starting rates for the third time in 2026: NVIDIA H100 SXM moved from $2.89 to $3.19 / HR (up 10%) and H200 from $3.49 to $3.99 / HR (up 14%), while B200 held at $5.99 / HR. H100 SXM has now risen in three consecutive captures — $1.50 (June) → $2.89 (July) → $3.19 (August) — under the same “refreshed weekly based on the best available rates from suppliers” mechanism.
The more notable shift is to the page’s own headline claim: the “Affordable compute” banner previously advertised GPUs “starting at $0.20/GPU/hr” with no bookable card anywhere near that price (only relative comparison multipliers for RTX 4090/3080/3070, no absolute rate). That banner now reads “starting at $3.19/GPU/hr” — exactly the H100 SXM rate — closing the gap between the advertised floor and the cheapest GPU actually listed. Serverless inference token pricing ($0.10–$4.00/M across the open-model catalog) is unchanged, and the per-instance price lock still means only new launches see each increase.
On-demand starting rates move again: H100 SXM $2.89 to $3.19/GPU-hr (+10.4%), H200 $3.49 to $3.99 (+14.3%); B200 holds at $5.99. The marketplace's 'starting at' claim, previously an unverifiable $0.20/GPU/hr floor with no matching card, now equals the H100 SXM rate exactly. Serverless inference rates and account-tier mechanics are unchanged.