grab▲gpu

GPU model · price + stock · hourly snapshot

Nvidia RTX 5060 cloud rental prices

3 providers list the Nvidia RTX 5060, 3 have it in stock. Prices run from $0.077 to $0.35 per GPU hour — the same card, 4.6x apart.

Blackwell 8 GB VRAM Q2 2025 5 configurations

LIVE Prices and stock checked 1 minute ago

Cheapest in stock
$0.077
Clore.ai
Median
$0.10
across 3 providers
Most expensive
$0.35
same card, 4.6x the price
You overpay by
27%
if you take the median instead of the floor

Providers

Every provider renting the Nvidia RTX 5060

All providers →

Billing:

On-demandfrom $0.077/hr3 providers in stockPay as you go, stop any time.Spotfrom $0.068/hr2 providers in stockCheaper, but the provider can take the machine back.Reservednone listedPrepay a term for the lowest hourly rate.
Providers renting the Nvidia RTX 5060, in stock first, then cheapest first
Provider Country Configs Scale From Stock Days in stock Grab
Clore.ai — 2 9x-9x $0.077 In stock 14/14 Grab
Vast.ai United States of America 2 1x-1x $0.10 per sec In stock 60/60 Grab
OctaSpace — 1 1x-1x $0.35 In stock 14/14 Grab

In stock first, then cheapest. Stock is what the provider reported at our last check. One GPU for a month (730h) at $0.077: $56.

Price history

Price trend

Cheapest Nvidia RTX 5060 rate has risen 35.0% since Sep 6 — $0.055 → $0.075/GPU/hr.

$0.075 /GPU/hr · Nov 4

One point per day: the lowest listed $/GPU/hr across every provider we track, captured daily. 54 offers in the latest snapshot.

Configurations

All 5 configurations

5 of 5 configurations
Nvidia RTX 5060 configurations from every provider
ConfigurationProviderRegionVRAMvCPURAMPer GPU/hrTotal/hrStock
9x NVIDIA GeForce RTX 5060SpotClore.aiClore.aiIN72 GB—16 GB$0.07$0.61In stock
1x RTX 5060SpotVast.aiVast.aiGB8 GB1223 GB$0.08per sec$0.08In stock
9x NVIDIA GeForce RTX 5060Clore.aiClore.aiIN72 GB—16 GB$0.08$0.69In stock
1x RTX 5060Vast.aiVast.aiIN8 GB1230 GB$0.10per sec$0.10In stock
1x NVIDIA GeForce RTX 5060OctaSpaceOctaSpaceDK8 GB2023 GB$0.35$0.35In stock

Model fit

What fits in 8 GB

The question behind most rentals: will the model load. Sized against one Nvidia RTX 5060.

VRAM required to serve open-weight LLMs on a Nvidia RTX 5060, by quantization precision
Model 16-bit8-bit4-bit
DeepSeek-V3 DeepSeek · 671B params · 37B active 1,610 GB 202× · 6 GB free 805 GB 101× · 3 GB free 403 GB 51× · 5 GB free
Llama 3.1 405B Meta · 405B params 972 GB 122× · 4 GB free 486 GB 61× · 2 GB free 243 GB 31× · 5 GB free
Qwen3 235B-A22B Alibaba · 235B params · 22B active 564 GB 71× · 4 GB free 282 GB 36× · 6 GB free 141 GB 18× · 3 GB free
Mixtral 8x22B Mistral · 141B params · 39B active 338 GB 43× · 6 GB free 169 GB 22× · 7 GB free 85 GB 11× · 3 GB free
Mistral Large 2 Mistral · 123B params 295 GB 37× · 1 GB free 148 GB 19× · 4 GB free 74 GB 10× · 6 GB free
gpt-oss-120b OpenAI · 117B params · 5.1B active 281 GB 36× · 7 GB free 140 GB 18× · 4 GB free 70 GB 9× · 2 GB free
Qwen2.5 72B Alibaba · 72.7B params 174 GB 22× · 2 GB free 87 GB 11× · 1 GB free 44 GB 6× · 4 GB free
Llama 3.3 70B Meta · 70.6B params 169 GB 22× · 7 GB free 85 GB 11× · 3 GB free 42 GB 6× · 6 GB free
Qwen3 32B Alibaba · 32.8B params 79 GB 10× · 1 GB free 39 GB 5× · 1 GB free 20 GB 3× · 4 GB free
Gemma 2 27B Google · 27.2B params 65 GB 9× · 7 GB free 33 GB 5× · 7 GB free 16 GB 3× · 8 GB free
Mistral Small 3 Mistral · 24B params 58 GB 8× · 6 GB free 29 GB 4× · 3 GB free 14 GB 2× · 2 GB free
Llama 3.1 8B Meta · 8B params 19 GB 3× · 5 GB free 10 GB 2× · 6 GB free 5 GB 1× · 3 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 5060.

Specifications

Nvidia RTX 5060 specifications

Nvidia RTX 5060 hardware specifications
ArchitectureBlackwell
Memory per GPU8 GB GDDR7
Memory bandwidth448 GB/s
Release dateQ2 2025
FP4 compute rate307 TFLOPS (dense), 614 with sparsity
FP8 compute rate76.8 TFLOPS (dense), 153.6 with sparsity
FP16 / BF16 compute rate38.4 TFLOPS (dense), 76.8 with sparsity
INT8 compute rate153.6 TOPS (dense), 307.2 with sparsity
Process nodeTSMC 4N
Board power145 W

Alternatives

Alternatives to the Nvidia RTX 5060