CanItRun Logocanitrun.

NVIDIA RTX 3080 10GB

The NVIDIA RTX 3080 10GB has 10 GB VRAM and 760 GB/s memory bandwidth. It can run 25 of our 70 tracked models natively in VRAM at 8k context.

With 10 GB GDDR6X, the NVIDIA RTX 3080 10GB is a consumer-tier GPU that can run 25 models natively. It's best for smaller models under 8B parameters.

NVIDIA RTX 3080 10GB: 10GB GDDR6X at 760 GB/s — bandwidth excellent, capacity limiting.

7B at Q4-Q5. 8B tight. ~8-14 t/s for 7B.

Full CUDA support. 10GB is awkward for LLM — consider 12GB or 16GB instead.

VendorNVIDIA
ArchitectureAmpere
VRAM10 GB
Memory typeGDDR6X
Memory bandwidth760 GB/s
Compute backendCUDA
TierConsumer
Released2020
Models (native)25 / 70
Models (offload)22 / 70
Software: Full llama.cpp and Ollama support out of the box. CUDA 12.x recommended; driver ≥ 525 required.

Popular models for this GPU

Models this GPU runs natively in VRAM (25)

Models that fit with CPU offload (22)

These use system RAM for layers that don't fit in VRAM — expect much slower inference.

Too large for this GPU (23)

Compare NVIDIA RTX 3080 10GB with other GPUs

Frequently asked questions

How much VRAM does the NVIDIA RTX 3080 10GB have?
The NVIDIA RTX 3080 10GB has 10 GB of GDDR6X with 760 GB/s memory bandwidth.
What is the NVIDIA RTX 3080 10GB best for?
With 10 GB of VRAM, the NVIDIA RTX 3080 10GB is best for running compact models (1B–8B) at low quantization, suitable for edge inference, prototyping, and lightweight tasks.
What LLMs can the NVIDIA RTX 3080 10GB run locally?
The NVIDIA RTX 3080 10GB can run 25 of the 70 open-weight models tracked by CanItRun natively in VRAM at 8k context. Top options include: Llama 3.1 8B Instruct at NVFP4, Llama 3.2 3B Instruct at BF16, Llama 3.2 1B Instruct at FP32.
Can the NVIDIA RTX 3080 10GB run Llama 3.3 70B Instruct?
The NVIDIA RTX 3080 10GB can run Llama 3.3 70B Instruct with CPU offload at Q2_K quantization, but inference will be slower than native VRAM execution.
Can the NVIDIA RTX 3080 10GB run Qwen 3.6 27B?
The NVIDIA RTX 3080 10GB can run Qwen 3.6 27B with CPU offload at NVFP4 quantization, but inference will be slower than native VRAM execution.
Can the NVIDIA RTX 3080 10GB run Llama 3.1 8B Instruct?
Yes. The NVIDIA RTX 3080 10GB runs Llama 3.1 8B Instruct natively in VRAM at NVFP4 quantization, achieving approximately 190 tokens per second.