How much VRAM does Llama 3.3 70B need?

Llama 3.3 70B (70B parameters) requires approximately 62.8 GB of memory with Q4_K_M quantization.

What is the best quantization for Llama 3.3 70B?

The recommended quantization for Llama 3.3 70B is Q4_K_M, which balances quality and memory efficiency.

Can it run?

Yes, NVIDIA H800 80GB can run Llama 3.3 70B with a B grade (Runs well). Expected decode speed: 50.6 tok/s.

BGood

Runs well

Using Q4_K_M in Ollama

Capabilities:

Fit status

Runs well

Decode

50.6 tok/s

TTFT

3827 ms

Safe context

20K

Memory

62.8 GB / 80.0 GB

Weights42.7 GB

KV Cache10.9 GB

Runtime1.2 GB

Headroom8.0 GB

Workload	Grade	Fit	Decode	TTFT	Context
Agentic Coding	C	Tight fit	56.9 tok/s	4948 ms	35K
Chat	B	Runs well	56.9 tok/s	1856 ms	11K
Coding	B	Runs well	50.6 tok/s	3827 ms	20K
RAG	C	Tight fit	56.9 tok/s	6185 ms	35K
Reasoning	B	Runs well	56.9 tok/s	4020 ms	20K

How Llama 3.3 70B (70B params) fits at each quantization level on NVIDIA H800 80GB (80.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	27.3 GB	Low	D37
Q3_K_S	3	34.3 GB	Low	D39
NVFP4	4

Ollama

ollama run llama-3.3-70b

HuggingFace

huggingface-cli download llama-3.3-70b

Upgrade options

NVIDIA GH200 96GBBudget pick

B75.9 tok/s decode

NVIDIA H20 96GBBiggest leap

B75.9 tok/s decode