Sign in Get started

GPU catalog

Choose your silicon

Every GPU can be configured with 32–124 GB of VRAM and RAM. Prices shown are the base hourly rate.

Popular

NVIDIAReady

H200 SXM

$1.00/ hr

141 GB

VRAM

128 GB

RAM

64

vCPU

Popular

NVIDIALow

RTX 5090

$0.69/ hr

32 GB

VRAM

31 GB

RAM

8

vCPU

NVIDIALow

H100 NVL

$2.59/ hr

94 GB

VRAM

150 GB

RAM

19

vCPU

NVIDIAReady

H100 SXM

$2.49/ hr

80 GB

VRAM

96 GB

RAM

14

vCPU

NVIDIAReady

H100 PCIe

$2.19/ hr

80 GB

VRAM

96 GB

RAM

14

vCPU

NVIDIAReady

A100 80GB

$1.89/ hr

80 GB

VRAM

96 GB

RAM

12

vCPU

NVIDIAReady

RTX PRO 6000

$1.49/ hr

96 GB

VRAM

64 GB

RAM

8

vCPU

NVIDIAReady

L40S

$0.99/ hr

48 GB

VRAM

96 GB

RAM

12

vCPU

NVIDIALow

L40

$0.69/ hr

48 GB

VRAM

151 GB

RAM

9

vCPU

NVIDIALow

RTX 6000 Ada

$0.74/ hr

48 GB

VRAM

109 GB

RAM

32

vCPU

NVIDIALow

RTX 5000 Ada

$0.49/ hr

32 GB

VRAM

62 GB

RAM

6

vCPU

NVIDIAReady

RTX 4090

$0.34/ hr

24 GB

VRAM

31 GB

RAM

12

vCPU

NVIDIAUnavailable

RTX 4000 Ada

$0.29/ hr

32 GB

VRAM

48 GB

RAM

8

vCPU

Sold out

NVIDIAUnavailable

RTX 4080 SUPER

$0.28/ hr

32 GB

VRAM

48 GB

RAM

8

vCPU

Sold out

NVIDIAUnavailable

RTX 5080

$0.49/ hr

32 GB

VRAM

48 GB

RAM

8

vCPU

Sold out

Model store

400+ open models

Run any open LLM.
If it fits in VRAM, it runs.

From DeepSeek to Qwen to Llama — deploy the latest open-weight models on DeployVolt in seconds. Bring your own weights, or pull a ready image. No lock-in, no per-token fees, just raw silicon at hourly rates.

deployvolt — launch

$ deploy --model deepseek-r1:671b --gpu h200 --vram 141gb

▸ Pulling image… ✓ done in 0.4s

▸ Server ready instantly — API at :8000

DS 671B MoE

DeepSeek V3.1

DeepSeek

128K ctx·H200 · 141GB

R1 671B MoE

DeepSeek R1

DeepSeek

64K ctx·H200 · 141GB

Q3 235B MoE

Qwen 3

Alibaba

128K ctx·H200 · 141GB

QC 32B

Qwen 3 Coder

Alibaba

128K ctx·H100 · 80GB

L4 400B MoE

Llama 4 Maverick

Meta

128K ctx·H200 · 141GB

L3 70B

Llama 3.3 70B

Meta

128K ctx·H100 · 80GB

ML 123B

Mistral Large 3

Mistral AI

128K ctx·H100 · 80GB

MS 24B

Mistral Small 3.1

Mistral AI

32K ctx·RTX 5090 · 32GB

GE 27B

Gemma 3 27B

Google

128K ctx·RTX 5090 · 32GB

PH 14B

Phi-4 14B

Microsoft

16K ctx·RTX 5090 · 32GB

+ your model400+ open-weight models.
Pick a GPU and go.Browse GPUs

vLLMSGLangOllamaTorchServeHugging Face