GPU NVIDIA

T4

Architecture · 16GB · PCIe

VRAM
16GB
FP16
-
TDP
-
Hardware Price
-
Cloud from
$0.350/hr
12 providers
Cheapest at Google Cloud Platform →

Quick Insights

Performance/Dollar
N/A
FP16 performance per $1000
VRAM/Dollar
N/A
VRAM per $1000
vs null Average
N/A
FP16 TFLOPS comparison
Cloud Availability
12 providers
from $0.350/hr

Specifications

VRAM 16GB
Memory Bandwidth -
FP16 TFLOPS -
Tensor TFLOPS -
FP32 TFLOPS -
TDP -
Form Factor -
Architecture -
NVLink No
Release Date -

Cloud GPU Pricing

Rent T4 from 12 cloud providers. Prices shown per GPU per hour.

Provider Type Instance GPUs On-Demand Per GPU Spot Availability
Google Cloud Platform hyperscaler gcp-t4 1x $0.350/hr $0.350/hr Cheapest $0.140/hr (-60%) -
Amazon Web Services hyperscaler g4dn.xlarge 1x $0.526/hr $0.526/hr $0.220/hr (-58%) -
Microsoft Azure hyperscaler Standard_NC4as_T4_v3 1x $0.526/hr $0.526/hr - -
Amazon Web Services hyperscaler g4dn.2xlarge 1x $0.752/hr $0.752/hr $0.306/hr (-59%) -
Microsoft Azure hyperscaler Standard_NC8as_T4_v3 1x $0.752/hr $0.752/hr - -
Amazon Web Services hyperscaler g4dn.12xlarge 4x $3.91/hr $0.978/hr $1.73/hr (-56%) -
Amazon Web Services hyperscaler g4dn.metal 8x $7.82/hr $0.978/hr $3.52/hr (-55%) -
Microsoft Azure hyperscaler Standard_NC64as_T4_v3 4x $4.35/hr $1.09/hr - -
Amazon Web Services hyperscaler g4dn.4xlarge 1x $1.20/hr $1.20/hr $0.526/hr (-56%) -
Microsoft Azure hyperscaler Standard_NC16as_T4_v3 1x $1.20/hr $1.20/hr - -
Amazon Web Services hyperscaler g4dn.8xlarge 1x $2.18/hr $2.18/hr $0.873/hr (-60%) -
Amazon Web Services hyperscaler g4dn.16xlarge 1x $4.35/hr $4.35/hr $1.56/hr (-64%) -
Best Spot Deal: Amazon Web Services offers spot pricing at $1.56/hr (64% off on-demand).

T4 vs Alternatives

Compare T4 with similar GPUs from other brands.

GPU VRAM FP16 TFLOPS Bandwidth Hardware Price Cloud Price
T4 Current 16GB - - - - -
AMD Radeon RX 7900 XT AMD 20GB (+25%) 104.0 800 GB/s - - Compare
AMD Radeon RX 7900 XTX AMD 24GB (+50%) 122.0 960 GB/s - - Compare
AMD Instinct MI100 AMD 32GB (+100%) 184.6 1.2 TB/s - - Compare

Best Use Cases

No specific use case recommendations for T4 yet.

Browse All Use Cases →

Compare T4

Other NVIDIA GPUs
Alternatives
Add to Comparison Tool

Frequently Asked Questions about T4

Pricing for T4 varies. Check our cloud pricing section for rental options starting at $0.350/hr.

Yes, the T4 with 16GB VRAM is suitable for many AI/ML workloads. For large language models, you may need multiple GPUs or consider higher-VRAM options like A100 or H100.

Consider buying for long-term, heavy usage (>4 hrs/day). Rent from cloud providers for short-term projects, experimentation, or when you need to scale quickly.

With 16GB VRAM and - FP16 TFLOPS, the T4 can run: Stable Diffusion, smaller LLMs (7B quantized), deep learning training, and gaming at high settings.

The T4 offers 16GB VRAM and - FP16 performance at its price point. Compare with similar GPUs using our comparison tool above. Key factors: VRAM for model size, TFLOPS for speed, and price for budget.