GPU model · price + stock · hourly snapshot
Nvidia RTX 4000 Ada cloud rental prices
3 providers list the Nvidia RTX 4000 Ada, 2 have it in stock. Prices run from $0.20 to $0.79 per GPU hour — the same card, 4.0x apart.
LIVE Prices and stock checked 1 minute ago
Providers
Every provider renting the Nvidia RTX 4000 Ada
| Provider | From | Stock | Grab |
|---|---|---|---|
| Runpod | $0.20 | In stock | Grab |
| Vast.ai | $0.20 | In stock | Grab |
| DigitalOcean | $0.79 | Out of stock | Grab |
In stock first, then cheapest. Stock is what the provider reported at our last check. One GPU for a month (730h) at $0.20: $146.
The Nvidia RTX 4000 Ada sells out fast. Get one email when it is back in stock.
Price history
Price trend
$0.200 /GPU/hr · Oct 22
One point per day: the lowest listed $/GPU/hr across every provider we track, captured daily. 5 offers in the latest snapshot.
Configurations
All 3 configurations
| Configuration | VRAM | Per GPU/hr | Stock |
|---|---|---|---|
| 1x RTX 4000 AdaRunpod | 20 GB | $0.20 | In stock |
| 1x RTX 4000AdaVast.ai | 20 GB | $0.20 | In stock |
| 1x RTX4000AdaDigitalOcean | 20 GB | $0.79 | Out of stock |
Model fit
What fits in 20 GB
The question behind most rentals: will the model load. Sized against one Nvidia RTX 4000 Ada.
| Model | 16-bit | 8-bit | 4-bit |
|---|---|---|---|
| DeepSeek-V3 DeepSeek · 671B params · 37B active | 1,610 GB 81× · 10 GB free | 805 GB 41× · 15 GB free | 403 GB 21× · 17 GB free |
| Llama 3.1 405B Meta · 405B params | 972 GB 49× · 8 GB free | 486 GB 25× · 14 GB free | 243 GB 13× · 17 GB free |
| Qwen3 235B-A22B Alibaba · 235B params · 22B active | 564 GB 29× · 16 GB free | 282 GB 15× · 18 GB free | 141 GB 8× · 19 GB free |
| Mixtral 8x22B Mistral · 141B params · 39B active | 338 GB 17× · 2 GB free | 169 GB 9× · 11 GB free | 85 GB 5× · 15 GB free |
| Mistral Large 2 Mistral · 123B params | 295 GB 15× · 5 GB free | 148 GB 8× · 12 GB free | 74 GB 4× · 6 GB free |
| gpt-oss-120b OpenAI · 117B params · 5.1B active | 281 GB 15× · 19 GB free | 140 GB 8× · 20 GB free | 70 GB 4× · 10 GB free |
| Qwen2.5 72B Alibaba · 72.7B params | 174 GB 9× · 6 GB free | 87 GB 5× · 13 GB free | 44 GB 3× · 16 GB free |
| Llama 3.3 70B Meta · 70.6B params | 169 GB 9× · 11 GB free | 85 GB 5× · 15 GB free | 42 GB 3× · 18 GB free |
| Qwen3 32B Alibaba · 32.8B params | 79 GB 4× · 1 GB free | 39 GB 2× · 1 GB free | 20 GB 1× · 0 GB free |
| Gemma 2 27B Google · 27.2B params | 65 GB 4× · 15 GB free | 33 GB 2× · 7 GB free | 16 GB 1× · 4 GB free |
| Mistral Small 3 Mistral · 24B params | 58 GB 3× · 2 GB free | 29 GB 2× · 11 GB free | 14 GB 1× · 6 GB free |
| Llama 3.1 8B Meta · 8B params | 19 GB 1× · 1 GB free | 10 GB 1× · 10 GB free | 5 GB 1× · 15 GB free |
Required memory is weights plus 20% for KV cache and runtime at short context; long contexts and large batches need more. Cells tinted green fit on a single Nvidia RTX 4000 Ada.
Specifications
Nvidia RTX 4000 Ada specifications
| Architecture | Ada Lovelace |
|---|---|
| Memory per GPU | 20 GB GDDR6 |
| Memory bandwidth | 360 GB/s |
| Release date | Q3 2023 |
| FP8 compute rate | 213.8 TFLOPS (dense), 427.6 with sparsity |
| FP16 / BF16 compute rate | 106.9 TFLOPS (dense), 213.8 with sparsity |
| Process node | TSMC 4N |
| Board power | 130 W |
Alternatives