Anthropic
Claude 5 Sonnet (175B)
High-quality balanced model, default on claude.ai
Model Summary
Family
Claude 5
Version
5.0
Parameters
175B (est.)
Parameter counts for closed models are estimates; vendors rarely publish exact sizes.
VRAM Requirements by Quantization
Memory needed to serve Claude 5 Sonnet (175B) for inference, including a 20% overhead for activations and KV cache.
| Precision | VRAM needed | Smallest single GPU that fits |
|---|---|---|
| INT4 (4-bit) | 97.79 GB | AMD Instinct MI250X (128 GB) |
| INT8 (8-bit) | 195.58 GB | AMD Instinct MI325X (256 GB) |
| FP16 (16-bit) | 391.16 GB | Multi-GPU required |
| FP32 (32-bit) | 782.31 GB | Multi-GPU required |
Recommended GPU Configurations
Cheapest on-demand configurations to serve Claude 5 Sonnet (175B) at 8-bit (196 GB VRAM).
2x AMD Instinct MI250X
256 GB total VRAM · CDNA 2
~$5.00/h
13x NVIDIA T4
208 GB total VRAM · Turing · multi-node
~$6.50/h
13x NVIDIA P100 SXM2
208 GB total VRAM · Pascal · multi-node
~$7.80/h
Quick GPU Planning
Use the calculator pre-filled with this exact version to estimate memory, speed, and compute requirements in a few clicks.
Access Pre-filled CalculatorFrequently Asked Questions
How much VRAM do you need to run Claude 5 Sonnet (175B)?
With an estimated 175B parameters, Claude 5 Sonnet (175B) needs roughly 196 GB of VRAM in 8-bit (INT8), 98 GB in 4-bit, and 391 GB in FP16, including a 20% overhead for activations and KV cache.
Which GPUs can run Claude 5 Sonnet (175B)?
At 8-bit quantization, the most cost-effective option is 2x AMD Instinct MI250X (256 GB combined VRAM, around $5.00/hour on-demand). Higher-end cards like the NVIDIA B200 or AMD MI355X reduce the GPU count needed.
Can Claude 5 Sonnet (175B) run on a single GPU?
Yes. In 8-bit, a single AMD Instinct MI325X (256 GB) fits the model.
How much does it cost to serve Claude 5 Sonnet (175B) in the cloud?
Renting 2x Instinct MI250X costs on the order of $5.00/hour, i.e. about $3,650/month running 24/7. Actual prices vary by provider and commitment; spot and reserved capacity can be significantly cheaper.
Deploy on a GPU cloud
Rent 2x Instinct MI250X by the hour instead of buying hardware.
