GPU cloud · head-to-head
Hyperstack vs RunPod
On the lowest published gpu cloud price in this table, Hyperstack is the cheaper of the two today. Prices change — every figure links to the provider’s own page, checked 16 Sept 2026.
- Hyperstack from
- $1.00/GPU-hr
- RunPod from
- $4.79/GPU-hr
- Plans compared
- 8
Hyperstack
GPU cloud · HQ United Kingdom
- Lowest published price
- $1.00/GPU-hr
- Plans listed
- 7
- With a published price
- 7
- Pricing
- —
- Data-centre countries
- —
- Certifications (sourced)
- SOC2, ISO27001
- API · Terraform · free tier
- —
RunPod
GPU cloud · HQ United States
- Lowest published price
- $4.79/GPU-hr
- Plans listed
- 1
- With a published price
- 1
- Pricing
- —
- Data-centre countries
- —
- Certifications (sourced)
- HIPAA, GDPR, ISO27001, SOC2, SOC3
- API · Terraform · free tier
- —
Hyperstack and RunPod gpu cloud plans
One table, both providers, sorted by published price — lowest first. Plans without a published price sort last.
| # | Provider | Offer | On-demand | Spot | Committed | As published (original currency) · source | Actions |
|---|---|---|---|---|---|---|---|
| 1 | Hyperstack | L40 x1 1× L40 | $1.00/GPU-hr | — | — | ||
| 2 | Hyperstack | A100 x1 1× A100 80GB | $1.60/GPU-hr | — | — | ||
| 3 | Hyperstack | RTX PRO 6000 x1 | $1.85/GPU-hr | — | — | ||
| 4 | Hyperstack | H100 x1 1× H100 | $2.50/GPU-hr | $2.00/GPU-hr | — | ||
| 5 | Hyperstack | H200 x1 1× H200 | $3.99/GPU-hr | — | — | ||
| 6 | RunPod | H100 serverless 1× H100 | $4.79/GPU-hr | — | — | ||
| 7 | Hyperstack | B200 x1 1× B200 | $6.00/GPU-hr | — | — | ||
| 8 | Hyperstack | B300 x1 1× B300 | $7.40/GPU-hr | — | — |
Company facts, side by side
Only facts with a public source. “Provider’s claim” means the provider states it and we have not checked it independently.
| Fact | Hyperstack | RunPod |
|---|---|---|
| Headquarters | UK (London) | USA (San Francisco) |
| Markets | Canada; USA; Norway; Europe | Global (31 regions claimed) |
| Billing | Per-minute billing, hourly rates; on-demand, reserved, spot | Per-second billing, hourly rates; Community + Secure Cloud; Serverless per-second |
| Reserved / committed pricing | Yes - reserved 'starting from' prices | Yes - long-term commitments / reserved clusters via sales |
| Spot / pre-emptible | Yes - Spot VMs (H100 PCIe, RTX Pro 6000, A100, L40, A6000) | — |
| Regions | CANADA-1 Quebec; US-1 Texas; NORWAY-1 Vestland | 31 global regions (list not published) |
| India regions | None | — |
| GPU models offered | B300; B200; H200 SXM; H100 SXM/NVLink/PCIe; RTX PRO 6000 SE; A100 SXM/NVLink/PCIe; L40; A6000; RTX A4000; GB200 (mentioned, no price) | B300; B200; H200; H200 NVL; H100 SXM/PCIe/NVL; A100 PCIe/SXM; RTX PRO 6000; RTX PRO 4500; L40S; L40; L4; RTX 5090; RTX 4090; A40; A6000; MI300X |
| Certifications | SOC 2 Type 2 (NexGen Cloud news + KB); Tier 3 data centres; ISO 27001 on roadmap ('early 2026') - unconfirmed | ISO/IEC 27001:2022 (valid to 26 Aug 2029); SOC 2 Type II; SOC 3; HIPAA program; GDPR program |
| In their words | — | 'The AI Developer Cloud' - Pods, Serverless, Clusters, Hub; 1M+ developers claimed |
| Serves | — | Developers; Startups; AI labs |
| Minimum commitment | — | None for on-demand pods/clusters ('no commitments') |
Not sure which fits your workload?
Tell us the shape and region you need. We pass the request to up to 3 matching providers, Hyperstack and RunPod included where they fit.