Side-by-side GPU comparison: specs, memory, compute performance, and live cloud pricing.
Verdict
T4 has 4 GB more VRAM, making it better suited for large models and long context windows. For compute-bound workloads like training, T4 delivers 7.7× higher FP16 throughput.
Specifications
| M4 (16 GB) | T4 | |
|---|---|---|
| VRAM | 12 GB | 16 GB |
| VRAM Type | LPDDR5X | GDDR6 |
| Memory Bandwidth | 0.1 TB/s | 0.3 TB/s |
| FP16 Performance | 8 TFLOPS | 65 TFLOPS |
| Manufacturer | Apple | NVIDIA |
| FP8 Support | No | No |
| FP4 Support | No | No |
Price / Performance
Based on cheapest single-GPU on-demand pricing. Lower $/TFLOP = better compute value; lower $/GB = better memory value.
| M4 (16 GB) | T4 | |
|---|---|---|
| $/hr (cheapest) | — | $0.35 |
| $/TFLOP (compute value) | — | $0.0054 |
| $/GB VRAM (memory value) | — | $0.0219 |
Cloud Pricing
Cheapest on-demand price per provider (single GPU).
M4 (16 GB)
No cloud pricing available.
T4
| Provider | On-demand | Spot | Rent |
|---|---|---|---|
| Google Cloud | $0.35/hr | $0.10/hr | |
| Microsoft Azure | $0.53/hr | $0.06/hr | |
| Amazon Web Services | $0.53/hr | $0.09/hr |
Model Compatibility
Models from the catalog that fit on each GPU, grouped by required precision.
M4 (16 GB) (1000 models)
T4 (1000 models)
You might also compare…
Pricing data refreshed hourly · Last updated June 1, 2026 · Browse all comparisons