Computed from the 257 plans we track across 11 providers. No survey, no estimates — this is the live database, aggregated.
| Type | Plans tracked | Cheapest | Median |
|---|---|---|---|
| vps | 134 | $2.42 | $21.89 |
| dedicated | 72 | $44.00 | $155.33 |
| gpu | 38 | $29.00 | $289.00 |
| web | 13 | $0.89 | $6.00 |
Providers with at least 3 priced plans listing RAM
| # | Provider | Median $/GB RAM | Sample size |
|---|---|---|---|
| 1 | OVHcloud | $0.53 | 6 plans |
| 2 | Host-Unlimited | $1.08 | 9 plans |
| 3 | Hetzner | $1.16 | 35 plans |
| 4 | LumaDock | $1.44 | 17 plans |
| 5 | InterServer | $1.50 | 25 plans |
| 6 | ServerSP | $1.56 | 15 plans |
| 7 | DatabaseMart | $2.13 | 76 plans |
| 8 | Natro | $2.27 | 20 plans |
| 9 | Mevspace | $3.25 | 30 plans |
| 10 | Kamatera | $4.75 | 9 plans |
| 11 | Liquid Web | $6.00 | 11 plans |
| Plan | GPU | $/GB VRAM | Monthly |
|---|---|---|---|
| Basic GPU Dedicated Server - K80DatabaseMart | Nvidia Tesla K80 · 24GB | $5.38 | $129.00/mo |
| Multi-GPU Dedicated Server - 3xRTX A5000DatabaseMart | 3 x Quadro RTX A5000 · 72GB ×3 | $6.31 | $454.35/mo |
| Multi-GPU Dedicated Server - 3xRTX A6000DatabaseMart | 3 x Nvidia Quadro RTX A6000 · 144GB ×3 | $8.33 | $1,199.00/mo |
| GEX44Hetzner | NVIDIA RTX 4000 SFF Ada Generation · 20GB | $10.25 | $205.00/mo |
| Basic GPU Dedicated Server - T1000DatabaseMart | Nvidia Quadro T1000 · 8GB | $10.41 | $83.30/mo |
| Professional GPU VPS - A4000DatabaseMart | Quadro RTX A4000 · 16GB | $11.19 | $179.00/mo |
| Multi-GPU Dedicated Server - 2xRTX A5000DatabaseMart | 2 x Quadro RTX A5000 · 48GB ×2 | $11.23 | $539.00/mo |
| Enterprise GPU Dedicated Server - RTX A6000DatabaseMart | Nvidia Quadro RTX A6000 · 48GB | $11.44 | $549.00/mo |
| Enterprise GPU Dedicated Server - A40DatabaseMart | Nvidia A40 · 48GB | $11.44 | $549.00/mo |
| Advanced GPU Dedicated Server - V100DatabaseMart | Nvidia V100 · 16GB | $11.56 | $185.00/mo |
Figures are computed at page-generation time from every active plan in the ComparisonHost database (257 plans at last refresh) and update daily. Medians are used instead of averages so a handful of very expensive dedicated or multi-GPU plans can't skew the picture. Prices are list prices in USD as published by providers; promotional and first-term discounts are not modelled. Want the raw rows? Every table links back to the underlying live plan pages.