How much VRAM does Llama 3.1 8B need?

Llama 3.1 8B (8B parameters) requires approximately 8.3 GB of memory with Q4_K_M quantization.

What is the best quantization for Llama 3.1 8B?

The recommended quantization for Llama 3.1 8B is Q4_K_M, which balances quality and memory efficiency.

Can it run?

Can RTX 3080 10GB run Llama 3.1 8B?

Q: Can RTX 3080 10GB run Llama 3.1 8B?

Yes, RTX 3080 10GB can run Llama 3.1 8B with a C grade (Tight fit). Expected decode speed: 118.4 tok/s.

CUsable

Tight fit

Using Q4_K_M in Ollama

Capabilities:

Fit status

Tight fit

Decode

118.4 tok/s

TTFT

1636 ms

Safe context

19K

Memory

8.3 GB / 10.0 GB

Memory breakdown

Weights4.9 GB

KV Cache1.3 GB

Runtime1.2 GB

Headroom1.0 GB

Performance by workload

Workload	Grade	Fit	Decode	TTFT	Context
Agentic Coding	C	Runs with offload	118.4 tok/s	2379 ms	33K
Chat	B	Runs well	118.4 tok/s	892 ms	10K
Coding	C	Tight fit	118.4 tok/s	1636 ms	19K
RAG	C	Runs with offload	118.4 tok/s	2974 ms	33K
Reasoning	C	Tight fit	118.4 tok/s	1933 ms	19K

Quantization options

How Llama 3.1 8B (8B params) fits at each quantization level on RTX 3080 10GB (10.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	3.1 GB	Low	D35
Q3_K_S	3	3.9 GB	Low	D37
NVFP4	4

Get started

Ollama

ollama run llama-3.1-8b

HuggingFace

huggingface-cli download llama-3.1-8b

Upgrade options

Hardware that runs Llama 3.1 8B well

Intel Arc B580 12GBBudget pick

C44.9 tok/s decode

~$249 MSRP

RTX 3060 12GBBest value

C48.7 tok/s decode

~$329 MSRP

RTX 3080 12GBBiggest leap

B142 tok/s decode

~$799 MSRP

RTX 3080 Ti 12GBNVIDIA upgrade

B138.3 tok/s decode

~$1,199 MSRP

See all results for RTX 3080 10GB See all hardware for Llama 3.1 8B