Entry-level Ampere GPU for basic CUDA development.
Weekly median price per GPU per hour · Get the data
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.
No providers match .
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 3050 has 8 GB of VRAM. Even at 4-bit, the weights of most modern AI models exceed the card's usable VRAM, and context adds KV cache on top. Below are some open-weight LLMs, with the estimated memory and GPUs each one needs.
| Model | Memory (INT4 / FP4) | RTX 3050s needed | Cost /hr | Cost /mo |
|---|---|---|---|---|
Qwen3.8-27B
27B
|
19 GB
|
3
|
–
|
–
|
Gemma 4 31B
30.7B
|
24 GB
|
4
|
–
|
–
|
GPT-OSS-120B
117B · 5.1B active
|
68 GB
|
10
|
–
|
–
|
GLM-5.3-Flash
320B · 18B active
|
178 GB
|
25
|
–
|
–
|
MiniMax-M3
428B · 23B active
|
241 GB
|
34
|
–
|
–
|
DeepSeek V4.1 Flash
552B · 16B active
|
306 GB
|
43
|
–
|
–
|
Kimi K3
2.8T · 104B active
|
1,545 GB
|
215
|
–
|
–
|
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. Pricing methodology.
Nvidia RTX 3050 · Per GPU
| Compute · dense | |
|---|---|
| FP16 / BF16 | 18.2 TFLOPS 36.4 with sparsity |
| INT8 | 72.8 TOPS 145.6 with sparsity |
| FP32 | 9.1 TFLOPS |
| Precision support | FP16BF16TF32FP32INT4INT8 |
| Memory | |
|---|---|
| Capacity | 8 GB GDDR6 |
| Bandwidth | 224 GB/s |
| Bus width | 128-bit |
| ECC | No |
| Silicon | |
|---|---|
| Architecture | Ampere |
| Process | Samsung 8N |
| Shader cores | 2,560 CUDA cores |
| Matrix cores | 80 Tensor cores |
| Compute units | 20 SMs |
| Fabric and host | |
|---|---|
| Host interface | PCIe 4.0 |
| Power | |
|---|---|
| Board power | 130 W |
| Cooling | Active |
Source: official Nvidia RTX 3050 datasheet.
The cheapest verified in-stock estimate is $50 per month on-demand, $22 spot.
GetDeploying currently tracks RTX 3050 configs from 1 provider. Salad has verified in-stock on-demand configs. See the full price comparison above for every provider and config.
More than one: even at 4-bit, the weights of most modern AI models exceed one RTX 3050's usable VRAM, and context adds KV cache on top. Larger models run across multiple GPUs: the model table above shows the estimated memory and GPU count for popular open-weight LLMs.
8GB GDDR6 with basic Ampere Tensor Core support. Entry-level GPU for learning CUDA programming.
Very limited VRAM and compute. Better suited for CUDA learning than inference or training.
Nvidia RTX 3060
12 GB GDDR6 · 61% more bandwidth than the RTX 3050
Nvidia RTX 4060
8 GB GDDR6 · 21% more bandwidth than the RTX 3050
Nvidia RTX 3060 Ti
8 GB GDDR6 · 2x the bandwidth of the RTX 3050
Last updated
