Top Ampere consumer GPU for local AI research and rendering.
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 RTX 3090 Ti has 24 GB of VRAM. In practice, that's enough memory for roughly 21B parameters at 4-bit or 6B 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) | RTX 3090 Tis needed | Cost /hr | Cost /mo |
|---|---|---|---|---|
Qwen3.8-27B
27B
|
19 GB
|
1
|
–
|
–
|
Gemma 4 31B
30.7B
|
24 GB
|
2
|
–
|
–
|
GPT-OSS-120B
117B · 5.1B active
|
68 GB
|
4
|
–
|
–
|
GLM-5.3-Flash
320B · 18B active
|
178 GB
|
9
|
–
|
–
|
MiniMax-M3
428B · 23B active
|
241 GB
|
12
|
–
|
–
|
DeepSeek V4.1 Flash
552B · 16B active
|
306 GB
|
15
|
–
|
–
|
Kimi K3
2.8T · 104B active
|
1,545 GB
|
72
|
–
|
–
|
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 RTX 3090 Ti · Per GPU
| Compute · dense | |
|---|---|
| FP16 / BF16 | 80 TFLOPS 160 with sparsity |
| INT8 | 320 TOPS 640 with sparsity |
| FP32 | 40 TFLOPS |
| Precision support | FP16BF16TF32FP32INT4INT8 |
| Memory | |
|---|---|
| Capacity | 24 GB GDDR6X |
| Bandwidth | 1,008 GB/s |
| Bus width | 384-bit |
| ECC | No |
| Silicon | |
|---|---|
| Architecture | Ampere |
| Process | Samsung 8N |
| Transistors | 28.3 billion |
| Shader cores | 10,752 CUDA cores |
| Matrix cores | 336 Tensor cores |
| Compute units | 84 SMs |
| Fabric and host | |
|---|---|
| Host interface | PCIe 4.0 |
| Power | |
|---|---|
| Board power | 450 W |
| Cooling | Active |
As of September 13, 2026, we track 12 configs from 3 providers. Prices are per GPU per hour.
| Billing type | Configs | Cheapest |
|---|---|---|
On-demand | 6 | $0.15 (in stock, Vast.ai) |
Reserved | 4 | $0.22 (12 mo, in stock, Vast.ai) |
Spot | 2 | $0.10 |
No median for on-demand, reserved and spot: we only show this when at least 3 providers list the GPU on that billing type.
The cheapest verified in-stock estimate is $108 per month on-demand, $158 reserved (12 mo), $158 spot.
GetDeploying currently tracks RTX 3090 Ti configs from 3 providers. Vast.ai and Salad have verified in-stock on-demand configs. See the full price comparison above for every provider and config.
One RTX 3090 Ti runs models up to roughly 21B parameters at 4-bit quantization or 6B 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.
24GB GDDR6X with Ampere Tensor Cores. Strong consumer GPU for AI research. Can handle 13B models at FP16.
Previous generation without FP8 support. High power draw (450W). Subject to Nvidia's consumer EULA, which may restrict data center use. The RTX 4090 offers better efficiency.
Last updated
