datacenter
Rent Nvidia T4.
Turing
16GB VRAM
70W
From $0.070/hr
Per hour
$0.070
Per day
$1.68
Per week
$11.76
Per month
$50
Provider spread
4 providers ·
up to 87% cheaper at the low end
Cheapest · $0.070/hr on Thunder Compute
Median $0.44/hr
Most expensive · $0.53/hr on Microsoft Azure
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 $0.070/hr cheapest provider rate
Own on-prem
Electricity per day
—
70W TDP
Rent breakeven on $2,000 MSRP
— days
· — months
AI models that fit
See all 11 →
Run these on this GPU.
- DeepSeek R1 Distill Qwen 1.5B 1× · fp16
- Whisper Medium 1× · fp16
- Whisper Small 1× · fp16
Cloud instances
See all 4 →
Hyperscaler bundles.
Pre-configured on 4 clouds — from $0.07/hr total
($0.070/hr per GPU).
FAQ
Frequently asked.
How is the Nvidia T4 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 T4 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 T4?
Depends on the model size. With 16GB of VRAM you can fine-tune 7B models with QLoRA or run inference on 7B–13B models comfortably.
Spot vs on-demand on the Nvidia T4 — 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 T4?
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 T4 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
$0.07/hr
Lowest $/hr per GPU
$0.070/hr
Providers
4
Instance shapes
4
| Provider | Instance | GPUs | vCPU | RAM | Disk | $/hr | $/hr per GPU | |
|---|---|---|---|---|---|---|---|---|
| tnr-t4-on-demand | 1× | — | — | — | $0.07/hr | $0.070/hr | ||
| n1+t4 | 1× | — | — | — | $0.35/hr | $0.35/hr | ||
| g4dn.xlarge | 1× | — | — | — | $0.53/hr | $0.53/hr | ||
| NC4as_T4_v3 | 1× | — | — | — | $0.53/hr | $0.53/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 T4 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 T4 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 T4?
Depends on the model size. With 16GB of VRAM you can fine-tune 7B models with QLoRA or run inference on 7B–13B models comfortably.
Spot vs on-demand on the Nvidia T4 — 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 T4?
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 T4 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
FP16
INT8
INT8
INT8
Models that run on this GPU.
GPU-count + quantization recommendations covering fine-tuning, inference, and run-it-yourself scenarios on the Nvidia T4.
DeepSeek R1 Distill Qwen 1.5B
by DeepSeek
Required GPUs
1× Nvidia T4
Total VRAM
16 GB
Parameters
2B
Context
128K tokens
Whisper Medium
by OpenAI
Required GPUs
1× Nvidia T4
Total VRAM
16 GB
Parameters
1B
Context
30 tokens
Whisper Small
by OpenAI
Required GPUs
1× Nvidia T4
Total VRAM
16 GB
Parameters
0B
Context
30 tokens
Whisper Base
by OpenAI
Required GPUs
1× Nvidia T4
Total VRAM
16 GB
Parameters
0B
Context
30 tokens
Tiny — fits on a Raspberry Pi-class GPU.
Whisper Tiny
by OpenAI
Required GPUs
1× Nvidia T4
Total VRAM
16 GB
Parameters
0B
Context
30 tokens
Llama 3.2 1B
by Meta AI
Required GPUs
1× Nvidia T4
Total VRAM
16 GB
Parameters
1B
Context
128K tokens
Phone / browser deployable.
Whisper Large v3
by OpenAI
Required GPUs
1× Nvidia T4
Total VRAM
16 GB
Parameters
2B
Context
30 tokens
Stable Diffusion 1.5
by Stability AI
Required GPUs
1× Nvidia T4
Total VRAM
16 GB
Parameters
1B
DeepSeek R1 Distill Qwen 7B
by DeepSeek
Required GPUs
1× Nvidia T4
Total VRAM
16 GB
Parameters
8B
Context
128K tokens
Slow but workable on 16GB T4.
Llama 3.1 8B
by Meta AI
Required GPUs
1× Nvidia T4
Total VRAM
16 GB
Parameters
8B
Context
128K tokens
Older T4s on a budget.
Mistral 7B v0.3
by Mistral AI
Required GPUs
1× Nvidia T4
Total VRAM
16 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 T4 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 T4 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 T4?
Depends on the model size. With 16GB of VRAM you can fine-tune 7B models with QLoRA or run inference on 7B–13B models comfortably.
Spot vs on-demand on the Nvidia T4 — 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 T4?
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 T4 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.