Purpose-built for virtual desktop (VDI) deployments.
Weekly median price per GPU per hour
By provider shows one card per company, placed where its best offer ranked. By configuration lists every offer.
What you can rent comes first: in stock, then waitlist, then not reported, then out of stock. Priced offers rank ahead of quote-only, and on-demand ahead of other billing types.
Within a group, five factors set the order:
We match the provider, its country, GPU model, form factor, billing type, availability and instance name. Matching is partial and case-insensitive. Everyday words work too, so "interruptible" finds spot and "sold out" finds out of stock.
Heads up: A provider's own page may quote a different figure, for example on tax, region or a promotion. A monthly-only plan shows a derived hourly rate. Verify before provisioning. More on how we price.
One A16 has 16 GB of VRAM. In practice, that's enough memory for roughly 8B parameters at 4-bit or 2B at 16-bit, assuming a 32K context. Below are some open-weight LLMs, with the estimated memory and GPUs each one needs.
| Model | Memory (INT4 / FP4) | A16s needed | Cost /hr | Cost /mo |
|---|---|---|---|---|
Qwen3.8-27B
27B
|
19 GB
|
2
|
$1.12
|
$806
|
Gemma 4 31B
30.7B
|
24 GB
|
2
|
$1.12
|
$806
|
GPT-OSS-120B
117B · 5.1B active
|
68 GB
|
8
|
$4.48
|
$3,226
|
GLM-5.3-Flash
320B · 18B active
|
178 GB
|
13
|
–
|
–
|
MiniMax-M3
428B · 23B active
|
241 GB
|
17
|
–
|
–
|
DeepSeek V4.1 Flash
552B · 16B active
|
306 GB
|
22
|
–
|
–
|
Kimi K3
2.8T · 104B active
|
1,545 GB
|
108
|
–
|
–
|
Estimates based on the median on-demand rate. Memory is weights plus FP16/BF16 KV cache at 32K context per request. GPU counts assume 90% of advertised VRAM is usable. No guarantee of runtime support, usable performance, or that a matching quantized build exists. How we estimate costs.
Nvidia A16 · Per GPU
| Compute · dense | |
|---|---|
| FP16 / BF16 | 17.9 TFLOPS 35.9 with sparsity |
| INT8 | 35.9 TOPS 71.8 with sparsity |
| FP32 | 4.5 TFLOPS |
| Precision support | FP16BF16TF32FP32INT4INT8 |
| Memory | |
|---|---|
| Capacity | 16 GB GDDR6 |
| Bandwidth | 200 GB/s |
| Bus width | 128-bit |
| ECC | Yes |
| Silicon | |
|---|---|
| Architecture | Ampere |
| Shader cores | 1,280 CUDA cores |
| Matrix cores | 40 Tensor cores |
| Compute units | 10 SMs |
| Fabric and host | |
|---|---|
| Host interface | PCIe 4.0 x16 |
| Power | |
|---|---|
| Cooling | Passive |
Source: official Nvidia A16 datasheet.
As of September 13, 2026, the median on-demand price is $0.56 per GPU per hour across 3 providers.
| Billing type | Configs | Median /GPU/hr | Cheapest |
|---|---|---|---|
On-demand | 9 | $0.56 | $0.47 (in stock, Vultr) |
The cheapest verified in-stock on-demand A16 tracked by GetDeploying is $0.47 per GPU per hour from Vultr.
At 720 hours per month, one A16 costs an estimated $403 at the median on-demand price. Cheapest verified in stock: $338 per month on-demand.
GetDeploying currently tracks A16 configs from 3 providers. The cheapest verified in-stock on-demand configs come from Vultr, Runcrate and Sesterce. See the full price comparison above for every provider and config.
One A16 runs models up to roughly 8B parameters at 4-bit quantization or 2B at 16-bit, assuming a 32K context. Larger models run across multiple GPUs: the model table above shows the estimated memory and GPU count for popular open-weight LLMs.
4x 16GB GPUs on a single card (64GB total). Designed for virtual desktop (VDI) and multi-user GPU sharing. Hardware video encode/decode support.
Built for VDI and multi-user GPU sharing. Individual GPU dies have limited compute, so better suited for remote desktops and lightweight graphics than AI workloads.
Last updated
