datacenter
租用 Nvidia L4.
Ada Lovelace
24GB VRAM
72W
From $0.18/hr
Per hour
$0.18
Per day
$4.32
Per week
$30.24
Per month
$130
Provider spread
5 providers ·
up to 78% cheaper at the low end
Cheapest · $0.18/hr on TensorDock
Median $0.61/hr
Most expensive · $0.81/hr on Amazon Web Services
Price history
Daily median across providers.
Loading...
Where to rent it
All providers carrying this GPU.
Buy vs rent
Should you rent or own?
Your usage
Rent on cloud
Per day
—
Per month
—
at $0.18/hr cheapest provider rate
Own on-prem
Electricity per day
—
72W TDP
Rent breakeven on $2,500 MSRP
— days
· — months
Workloads
Suitable workloads.
AI models that fit
See all 7 →
Run these on this GPU.
- Whisper Large v3 1× · fp16
- Whisper Medium 1× · fp16
- DeepSeek R1 Distill Qwen 7B 1× · fp16
Cloud instances
See all 2 →
Hyperscaler bundles.
Pre-configured on 2 clouds — from $0.61/hr total
($0.61/hr per GPU).
FAQ
Frequently asked.
How is the Nvidia L4 price calculated?
We pull live listings from each provider's public API, take the median hourly rate across active offers, and refresh every hour. The rate shown is the median, so a single low-ball spot offer can't distort the headline.
Why does the Nvidia L4 cost different amounts on different providers?
P2P marketplaces like Vast.ai aggregate offers from individual hosts who set their own rates — bidding pushes prices down. First-party clouds (Lambda, hyperscalers) charge a managed-service premium for support, SLAs, and integrated networking. Decentralized networks (io.net, Akash) settle in tokens, which adds volatility but often the lowest base rate.
Can I really train an LLM on a single Nvidia L4?
Depends on the model size. With 24GB of VRAM you can fine-tune 13B–34B models with LoRA, or run inference on 70B models at int4 quantization.
Spot vs on-demand on the Nvidia L4 — which should I rent?
On-demand keeps the same instance until you stop it; spot (or interruptible) is cheaper but the host can reclaim it when a higher-paying job lands. Use on-demand for training runs and anything stateful. Use spot for stateless inference, batch jobs, and experiments where a checkpoint every few minutes is enough to recover.
How is hourly billing measured for the Nvidia L4?
Most providers bill per-second once the instance is running, with a small minimum (often 60 seconds). A handful of first-party clouds round up to the minute. Either way, headline $/hr is the right comparison unit.
Does the region of the host affect the Nvidia L4 price?
Yes — US and EU regions usually carry a premium over LATAM, India, and parts of APAC, especially on first-party clouds. P2P marketplaces hide this behind one global price because supply moves wherever bids exist.
Cloud instance options
Pre-configured instances on hyperscalers.
Whole-instance bundles (GPU + vCPU + RAM + disk) on the major clouds. Per-GPU rate often drops as the count rises. View = spec page · Launch = sign up (affiliate).
Cheapest bundle
$0.61/hr
Lowest $/hr per GPU
$0.61/hr
Providers
2
Instance shapes
2
| Provider | Instance | GPUs | vCPU | RAM | Disk | $/hr | $/hr per GPU | |
|---|---|---|---|---|---|---|---|---|
| g2-standard-4 | 1× | — | — | — | $0.61/hr | $0.61/hr | ||
| g6.xlarge | 1× | — | — | — | $0.81/hr | $0.81/hr |
Looking for the cheapest rate?
Hyperscaler bundles include managed networking + SLAs. Raw per-GPU rental on P2P marketplaces is typically 3–10× cheaper.
See raw rental rates on the Overview tab →
FAQ
Cloud instances — common questions.
How is the Nvidia L4 price calculated?
We pull live listings from each provider's public API, take the median hourly rate across active offers, and refresh every hour. The rate shown is the median, so a single low-ball spot offer can't distort the headline.
Why does the Nvidia L4 cost different amounts on different providers?
P2P marketplaces like Vast.ai aggregate offers from individual hosts who set their own rates — bidding pushes prices down. First-party clouds (Lambda, hyperscalers) charge a managed-service premium for support, SLAs, and integrated networking. Decentralized networks (io.net, Akash) settle in tokens, which adds volatility but often the lowest base rate.
Can I really train an LLM on a single Nvidia L4?
Depends on the model size. With 24GB of VRAM you can fine-tune 13B–34B models with LoRA, or run inference on 70B models at int4 quantization.
Spot vs on-demand on the Nvidia L4 — which should I rent?
On-demand keeps the same instance until you stop it; spot (or interruptible) is cheaper but the host can reclaim it when a higher-paying job lands. Use on-demand for training runs and anything stateful. Use spot for stateless inference, batch jobs, and experiments where a checkpoint every few minutes is enough to recover.
How is hourly billing measured for the Nvidia L4?
Most providers bill per-second once the instance is running, with a small minimum (often 60 seconds). A handful of first-party clouds round up to the minute. Either way, headline $/hr is the right comparison unit.
Does the region of the host affect the Nvidia L4 price?
Yes — US and EU regions usually carry a premium over LATAM, India, and parts of APAC, especially on first-party clouds. P2P marketplaces hide this behind one global price because supply moves wherever bids exist.
AI models
FP16
FP16
FP16
FP16
FP16
FP16
FP16
Models that run on this GPU.
GPU-count + quantization recommendations covering fine-tuning, inference, and run-it-yourself scenarios on the Nvidia L4.
Whisper Large v3
by OpenAI
Required GPUs
1× Nvidia L4
Total VRAM
24 GB
Parameters
2B
Context
30 tokens
Datacenter STT workhorse.
Whisper Medium
by OpenAI
Required GPUs
1× Nvidia L4
Total VRAM
24 GB
Parameters
1B
Context
30 tokens
DeepSeek R1 Distill Qwen 7B
by DeepSeek
Required GPUs
1× Nvidia L4
Total VRAM
24 GB
Parameters
8B
Context
128K tokens
Llama 3.2 3B
by Meta AI
Required GPUs
1× Nvidia L4
Total VRAM
24 GB
Parameters
3B
Context
128K tokens
Stable Diffusion XL
by Stability AI
Required GPUs
1× Nvidia L4
Total VRAM
24 GB
Parameters
4B
Llama 3.1 8B
by Meta AI
Required GPUs
1× Nvidia L4
Total VRAM
24 GB
Parameters
8B
Context
128K tokens
Mistral 7B v0.3
by Mistral AI
Required GPUs
1× Nvidia L4
Total VRAM
24 GB
Parameters
7B
Context
33K tokens
Renting for inference?
Pair these models with the cheapest provider on the rental table.
See rental rates on the Overview tab →
FAQ
AI models on this GPU.
How is the Nvidia L4 price calculated?
We pull live listings from each provider's public API, take the median hourly rate across active offers, and refresh every hour. The rate shown is the median, so a single low-ball spot offer can't distort the headline.
Why does the Nvidia L4 cost different amounts on different providers?
P2P marketplaces like Vast.ai aggregate offers from individual hosts who set their own rates — bidding pushes prices down. First-party clouds (Lambda, hyperscalers) charge a managed-service premium for support, SLAs, and integrated networking. Decentralized networks (io.net, Akash) settle in tokens, which adds volatility but often the lowest base rate.
Can I really train an LLM on a single Nvidia L4?
Depends on the model size. With 24GB of VRAM you can fine-tune 13B–34B models with LoRA, or run inference on 70B models at int4 quantization.
Spot vs on-demand on the Nvidia L4 — which should I rent?
On-demand keeps the same instance until you stop it; spot (or interruptible) is cheaper but the host can reclaim it when a higher-paying job lands. Use on-demand for training runs and anything stateful. Use spot for stateless inference, batch jobs, and experiments where a checkpoint every few minutes is enough to recover.
How is hourly billing measured for the Nvidia L4?
Most providers bill per-second once the instance is running, with a small minimum (often 60 seconds). A handful of first-party clouds round up to the minute. Either way, headline $/hr is the right comparison unit.
Does the region of the host affect the Nvidia L4 price?
Yes — US and EU regions usually carry a premium over LATAM, India, and parts of APAC, especially on first-party clouds. P2P marketplaces hide this behind one global price because supply moves wherever bids exist.