GPU model · price + stock · hourly snapshot
Nvidia RTX 3080 cloud rental prices
5 providers list the Nvidia RTX 3080, 4 have it in stock. Prices run from $0.092 to $0.22 per GPU hour — the same card, 2.4x apart.
LIVE Prices and stock checked 1 minute ago
Providers
Every provider renting the Nvidia RTX 3080
| Provider | From | Stock | Grab |
|---|---|---|---|
| Clore.ai | $0.092 | In stock | Grab |
| Vast.ai | $0.11 per sec | In stock | Grab |
| SimplePod | $0.12 | In stock | Grab |
| OctaSpace | $0.22 | In stock | Grab |
| Runpod | $0.17 per sec | Not reported | Grab |
In stock first, then cheapest. Stock is what the provider reported at our last check. One GPU for a month (730h) at $0.092: $67.
Price history
Price trend
$0.092 /GPU/hr · Nov 4
One point per day: the lowest listed $/GPU/hr across every provider we track, captured daily. 90 offers in the latest snapshot.
Configurations
All 7 configurations
| Configuration | VRAM | Per GPU/hr | Stock |
|---|---|---|---|
| 1x RTX 3080SpotVast.ai | 10 GB | $0.08per sec | In stock |
| 4x NVIDIA GeForce RTX 3080Clore.ai | 40 GB | $0.09 | In stock |
| 4x NVIDIA GeForce RTX 3080SpotClore.ai | 40 GB | $0.09 | In stock |
| 1x RTX 3080Vast.ai | 10 GB | $0.11per sec | In stock |
| 1x RTX 3080SimplePod | 10 GB | $0.12 | In stock |
| 1x NVIDIA GeForce RTX 3080OctaSpace | 10 GB | $0.22 | In stock |
| 1x RTX 3080Runpod | 10 GB | $0.17per sec | Not reported |
Model fit
What fits in 10 GB
The question behind most rentals: will the model load. Sized against one Nvidia RTX 3080.
| Model | 16-bit | 8-bit | 4-bit |
|---|---|---|---|
| DeepSeek-V3 DeepSeek · 671B params · 37B active | 1,610 GB 162× · 10 GB free | 805 GB 81× · 5 GB free | 403 GB 41× · 7 GB free |
| Llama 3.1 405B Meta · 405B params | 972 GB 98× · 8 GB free | 486 GB 49× · 4 GB free | 243 GB 25× · 7 GB free |
| Qwen3 235B-A22B Alibaba · 235B params · 22B active | 564 GB 57× · 6 GB free | 282 GB 29× · 8 GB free | 141 GB 15× · 9 GB free |
| Mixtral 8x22B Mistral · 141B params · 39B active | 338 GB 34× · 2 GB free | 169 GB 17× · 1 GB free | 85 GB 9× · 5 GB free |
| Mistral Large 2 Mistral · 123B params | 295 GB 30× · 5 GB free | 148 GB 15× · 2 GB free | 74 GB 8× · 6 GB free |
| gpt-oss-120b OpenAI · 117B params · 5.1B active | 281 GB 29× · 9 GB free | 140 GB 15× · 10 GB free | 70 GB 8× · 10 GB free |
| Qwen2.5 72B Alibaba · 72.7B params | 174 GB 18× · 6 GB free | 87 GB 9× · 3 GB free | 44 GB 5× · 6 GB free |
| Llama 3.3 70B Meta · 70.6B params | 169 GB 17× · 1 GB free | 85 GB 9× · 5 GB free | 42 GB 5× · 8 GB free |
| Qwen3 32B Alibaba · 32.8B params | 79 GB 8× · 1 GB free | 39 GB 4× · 1 GB free | 20 GB 2× · 0 GB free |
| Gemma 2 27B Google · 27.2B params | 65 GB 7× · 5 GB free | 33 GB 4× · 7 GB free | 16 GB 2× · 4 GB free |
| Mistral Small 3 Mistral · 24B params | 58 GB 6× · 2 GB free | 29 GB 3× · 1 GB free | 14 GB 2× · 6 GB free |
| Llama 3.1 8B Meta · 8B params | 19 GB 2× · 1 GB free | 10 GB 1× · 0 GB free | 5 GB 1× · 5 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 3080.
Specifications
Nvidia RTX 3080 specifications
| Architecture | Ampere |
|---|---|
| Memory per GPU | 10 GB GDDR6X |
| Memory bandwidth | 760 GB/s |
| Release date | Q3 2020 |
| FP16 / BF16 compute rate | 59.5 TFLOPS (dense), 119 with sparsity |
| INT8 compute rate | 238 TOPS (dense), 476 with sparsity |
| Process node | Samsung 8N |
| Board power | 320 W |
Alternatives