Home / Servers
Find a server for your AI model
Memory-capable means the selected model, quantization and context fit the server's GPU memory (and system RAM) using the same calculators as local hardware. Serving numbers, when shown, are official vLLM nightly measurements for matching GPU identity — not Phase 5 single-user tok/s and not this exact rental host.
Compare buy vs rent using these rental prices against a current local build or workstation.
Serving evaluation uses measured vLLM TTFT and TPOT at an explicit input/output shape and concurrency. Missing evidence is reported as insufficient, not as a fail.
Workload shapes are only those present in official vLLM nightly evidence. Request rate is not interchangeable with concurrency. P95 is not published by this source. Targets are yours, not an industry SLA.
Memory filter: Phi-3 Mini 128K Instruct at Q4 / 8K. Offers that are memory-capable remain listed even when serving evidence is absent.
| Configuration | Provider | Location | Price | Stock | |
|---|---|---|---|---|---|
| 1× NVIDIA RTX 5070 12 GB 31 GB RAM · 697 GB SSD | Vast.ai | VN | $0.074/hour · ~$54/month equivalent (hourly × 730) | Listed as available | Provider |
| 1× NVIDIA RTX 5080 16 GB 31 GB RAM · 677 GB SSD | Vast.ai | CA | $0.147/hour · ~$108/month equivalent (hourly × 730) | Listed as available | Provider |
| 1× NVIDIA RTX 4090 24 GB 23 GB RAM · 171 GB SSD | Vast.ai | US | $0.269/hour · ~$196/month equivalent (hourly × 730) | Listed as available | Provider |
| 2× NVIDIA RTX 5070 Ti 16 GB 62 GB RAM · 1304 GB SSD | Vast.ai | PL | $0.309/hour · ~$225/month equivalent (hourly × 730) | Listed as available | Provider |
| 4× NVIDIA RTX 3090 24 GB 252 GB RAM · 1280 GB SSD | Vast.ai | JP | $0.406/hour · ~$296/month equivalent (hourly × 730) | Listed as available | Provider |
| 2× NVIDIA RTX 5080 16 GB 126 GB RAM · 1555 GB SSD | Vast.ai | US | $0.430/hour · ~$314/month equivalent (hourly × 730) | Listed as available | Provider |
| 1× NVIDIA RTX PRO 6000 Blackwell 96 GB 188 GB RAM · 1879 GB SSD | Vast.ai | US | $1.069/hour · ~$780/month equivalent (hourly × 730) | Listed as available | Provider |
| 2× NVIDIA H100 NVL 94 GB 503 GB RAM · 4930 GB SSD | Vast.ai | US | $5.335/hour · ~$3894/month equivalent (hourly × 730) | Listed as available | Provider |
Priced, stock not confirmed as available
Shown separately. Not ranked as currently available.
| Configuration | Provider | Location | Price | Stock | |
|---|---|---|---|---|---|
| 2× NVIDIA L4 24 GB 192 GB RAM · ? NVME | OVHcloud | EU | $1204/month · $1204 setup | Stock not confirmed by provider | Provider |
| 2× NVIDIA L4 24 GB 192 GB RAM · ? NVME | OVHcloud | EU | $1241/month · $1241 setup | Stock not confirmed by provider | Provider |
| 2× NVIDIA L4 24 GB 192 GB RAM · ? NVME | OVHcloud | EU | $1279/month · $1279 setup | Stock not confirmed by provider | Provider |
| 4× NVIDIA L4 24 GB 192 GB RAM · ? NVME | OVHcloud | EU | $1577/month · $1204 setup | Stock not confirmed by provider | Provider |
| 4× NVIDIA L4 24 GB 192 GB RAM · ? NVME | OVHcloud | EU | $1614/month · $1241 setup | Stock not confirmed by provider | Provider |
| 4× NVIDIA L4 24 GB 192 GB RAM · ? NVME | OVHcloud | EU | $1652/month · $1279 setup | Stock not confirmed by provider | Provider |
| 1× NVIDIA GH200 96 GB 480 GB RAM · 960 GB NVME | Vultr | — | $2009/month · $0 setup | Not listed as available | Provider |
| 4× NVIDIA Tesla V100S 32 GB 192 GB RAM · ? NVME | OVHcloud | US | $3299/month · $3299 setup | Stock not confirmed by provider | Provider |
| 2× NVIDIA L40S 48 GB 384 GB RAM · ? NVME | OVHcloud | US | $4355/month · $4355 setup | Stock not confirmed by provider | Provider |
| 2× NVIDIA L40S 48 GB 384 GB RAM · ? NVME | OVHcloud | CA | $4355/month · $4355 setup | Stock not confirmed by provider | Provider |
| 2× NVIDIA L40S 48 GB 384 GB RAM · ? NVME | OVHcloud | EU | $4355/month · $4355 setup | Stock not confirmed by provider | Provider |
| 4× NVIDIA L40S 48 GB 384 GB RAM · ? NVME | OVHcloud | US | $5355/month · $4355 setup | Stock not confirmed by provider | Provider |
| 4× NVIDIA L40S 48 GB 384 GB RAM · ? NVME | OVHcloud | EU | $5355/month · $4355 setup | Stock not confirmed by provider | Provider |
| 4× NVIDIA L40S 48 GB 384 GB RAM · ? NVME | OVHcloud | CA | $5355/month · $4355 setup | Stock not confirmed by provider | Provider |
| 4× NVIDIA A100 80 GB 1024 GB RAM · 2 × 450 GB SSD | Vultr | — | $7000/month · $0 setup | Not listed as available | Provider |
Providers
GPU / dedicated AI compute only. Not a generic hosting directory. All providers with inventory
Want to own the hardware? Build a PC · Prefer a prebuilt? Workstations