Check my GPU

LLMs ·

Llama 3 8B Instruct

Parameters
8.0B
Architecture
Dense
Max context
8K tokens
License
llama3
Released
2024-04

How to run it

Quickest · Ollama

One command downloads a ready-to-run version and starts a chat.

ollama run llama3:8b

Get Ollama at ollama.com/download (opens in a new tab) · model page (opens in a new tab)

Quantized files · GGUF

Works in LM Studio, llama.cpp, Jan and Ollama. No sign-in needed.

huggingface.co/lmstudio-community/Meta-Llama-3-8B-Instruct-GGUF (opens in a new tab)

ollama run hf.co/lmstudio-community/Meta-Llama-3-8B-Instruct-GGUF

Original weights · Hugging Face

Full-precision files from Meta, for fine-tuning, vLLM, Transformers.

huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct (opens in a new tab)

Free Hugging Face account needed: request access on the model page, and Meta approves it (usually within minutes to a few hours).

Official repo · GitHub

Reference code, inference scripts and docs from Meta. The weights themselves download from Hugging Face.

github.com/meta-llama/llama3 (opens in a new tab)

git clone https://github.com/meta-llama/llama3

Meta’s official Llama 3 repo.

Links checked 6 days ago against Hugging Face, Ollama and GitHub.

Memory needed

Weights plus KV cache and runtime overhead. Leave a little extra free for the operating system and other apps.

Quantization Quality Weights Total at 8K context Download
Q2_K Low 3.2 GB 4.9 GB
Q3_K_M Low 3.9 GB 5.6 GB GGUF (IQ3_M) (opens in a new tab)
Q4_K_M Good 4.9 GB 6.6 GB GGUF (opens in a new tab)
Q5_K_M High 5.7 GB 7.4 GB GGUF (opens in a new tab)
Q6_K High 6.6 GB 8.3 GB GGUF (opens in a new tab)
Q8_0 Very high 8.5 GB 10.2 GB GGUF (opens in a new tab)
F16 Full 16.1 GB 17.7 GB Original weights (opens in a new tab)

Best hardware for Llama 3 8B Instruct

The least expensive GPU in our catalog that runs it without offloading is the Intel Arc A380 (launch price $139) at Q3_K_M.

Grade · points Hardware Quant Memory Speed Fit
73 ptsGreat (A) · 73 of 100 points: Model quality 18.9/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 15.0/15 AMD Radeon RX 9070 XT Q8_0 11.0 GB / 16.0 GB 42.9 tok/s Runs well
73 ptsGreat (A) · 73 of 100 points: Model quality 18.9/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 15.0/15 NVIDIA GeForce RTX 4070 Ti SUPER Q8_0 11.0 GB / 16.0 GB 52.3 tok/s Runs well
73 ptsGreat (A) · 73 of 100 points: Model quality 18.9/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 15.0/15 NVIDIA GeForce RTX 4080 SUPER Q8_0 11.0 GB / 16.0 GB 57.3 tok/s Runs well
73 ptsGreat (A) · 73 of 100 points: Model quality 18.9/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 15.0/15 NVIDIA GeForce RTX 5070 Ti Q8_0 11.0 GB / 16.0 GB 69.7 tok/s Runs well
73 ptsGreat (A) · 73 of 100 points: Model quality 18.9/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 15.0/15 NVIDIA GeForce RTX 5080 Q8_0 11.0 GB / 16.0 GB 74.7 tok/s Runs well
73 ptsGreat (A) · 73 of 100 points: Model quality 18.7/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 15.0/15 NVIDIA GeForce RTX 4070 SUPER Q6_K 8.9 GB / 12.0 GB 53.5 tok/s Runs well
73 ptsGreat (A) · 73 of 100 points: Model quality 18.7/45, Memory fit 25.0/25, Speed 14.7/15, Memory use 15.0/15 NVIDIA GeForce RTX 3060 12GB Q6_K 8.9 GB / 12.0 GB 38.2 tok/s Runs well
73 ptsGreat (A) · 73 of 100 points: Model quality 18.3/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 15.0/15 Intel Arc B580 Q5_K_M 8.0 GB / 12.0 GB 44.0 tok/s Runs well
72 ptsGreat (A) · 72 of 100 points: Model quality 18.9/45, Memory fit 25.0/25, Speed 15.0/15, Memory use 13.9/15 NVIDIA GeForce RTX 5090 F16 19.3 GB / 32.0 GB 64.3 tok/s Runs well
71 ptsGood (B) · 71 of 100 points: Model quality 18.0/45, Memory fit 24.0/25, Speed 15.0/15, Memory use 15.0/15 NVIDIA GeForce RTX 4060 Q4_K_M 7.0 GB / 8.0 GB 41.6 tok/s Tight fit

Check it on your own hardware

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.