Llama 2 13B Chat on AMD Ryzen AI Max+ 395 (96 GB)

Can the AMD Ryzen AI Max+ 395 (96 GB) run Llama 2 13B Chat?

Yes, it runs well

Runs well. At Q3_K_M it needs about 16.1 GB of 72.0 GB usable, with an estimated speed of 26.9 tok/s.

Best quant: Q3_K_M
Generation speed: 26.9 tok/s
Prompt processing: 2,829 tok/s
First token (1K prompt): 0.4 s
Largest safe context: 4K

Best app to run it on the AMD Ryzen AI Max+ 395 (96 GB)

On Windows

Lemonade

Desktop app · runs on llama.cpp, ONNX Runtime GenAI (Ryzen AI NPU)

Built for Ryzen AI and Strix Halo, including NPU + iGPU modes on Windows.

lemonade pull TheBloke/Llama-2-13B-chat-GGUF:Q3_K_M

LM Studio Point-and-click: search, download and chat in one window.
Ollama One command with Ollama’s own build of this model.

On Linux

Lemonade

Desktop app · runs on llama.cpp, ONNX Runtime GenAI (Ryzen AI NPU)

Built for Ryzen AI and Strix Halo, including NPU + iGPU modes on Windows.

lemonade pull TheBloke/Llama-2-13B-chat-GGUF:Q3_K_M

LM Studio Point-and-click: search, download and chat in one window.
Ollama One command with Ollama’s own build of this model.

How we choose: the files published for this model, your operating system and GPU, whether it fits fully in memory, and published speed tests (MLX vs llama.cpp on Apple silicon, vLLM vs llama.cpp vs Ollama on NVIDIA). Details.

Download Llama 2 13B Chat

Recommended for the AMD Ryzen AI Max+ 395 (96 GB): Q3_K_M. Every other quantization has its own download link in the table below.

Quickest · Ollama

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

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

ollama run llama2:13b-chat-q3_K_M

Quantized files · GGUF

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

Download the Q3_K_M file (opens in a new tab)

huggingface.co/TheBloke/Llama-2-13B-chat-GGUF (opens in a new tab)

ollama run hf.co/TheBloke/Llama-2-13B-chat-GGUF:Q3_K_M

Original weights · Hugging Face

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

huggingface.co/meta-llama/Llama-2-13b-chat-hf (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/llama (opens in a new tab)

Meta’s original Llama 2 repo (now archived).

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

Links checked 3 days ago against Hugging Face, Ollama and GitHub.

How it’s rated

Quality 21Basic quality (D) · 21 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking.: How capable the model itself is at Q3_K_M, from its editorial quality tier scaled by the quantization or precision. It’s the same on any hardware that runs this variant. Scale: A+ 85+ · A 70+ · B 55+ · C 40+ · D under 40.
Runs Runs well: How well it runs on the AMD Ryzen AI Max+ 395 (96 GB): speed, capped by how well it fits in memory. Runs great: 30+ tokens/s (or near the image, video or decision speed target) · Runs well: 10+ tokens/s · Slow: 3+ tokens/s · Very slow: under 3 · Won’t run: doesn’t fit even with offload. A tight fit or RAM offload caps it at Runs well; heavy offload caps it at Slow.

Match score

To order lists, each result also gets a match score out of 100 from four parts. It favors the most capable model that still runs well.

Model qualityEditorial quality tier of the model, scaled by the quantization or precision used.	9.8 / 45
Memory fitFull marks when the model fits in the share of unified memory the GPU can use; fewer when it goes past that limit.	25.0 / 25
SpeedGeneration speed against a 40 tokens/s target for LLMs, or render time against a target for image and video.	12.3 / 15
Memory useRewards using the memory you have, so a bigger, better variant beats a tiny one that leaves most memory idle.	5.2 / 15
Match score	52.3 / 100

The match score orders results; it is not shown as a grade. Quality and Runs are the two ratings in the tables.

Every quantization on the AMD Ryzen AI Max+ 395 (96 GB)

Quant	Memory needed	Available	Speed	Runs	Quality	Download
`Q2_K`	14.9 GB	72.0 GB	34.1 tok/s	Runs great	18Basic quality (D) · 18 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q3_K_M`	16.1 GB	72.0 GB	26.9 tok/s	Runs well	21Basic quality (D) · 21 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q4_K_M`	17.7 GB	72.0 GB	20.8 tok/s	Runs well	24Basic quality (D) · 24 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q5_K_M`	19.0 GB	72.0 GB	17.2 tok/s	Runs well	25Basic quality (D) · 25 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q6_K`	20.4 GB	72.0 GB	14.4 tok/s	Runs well	25Basic quality (D) · 25 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`Q8_0`	23.5 GB	72.0 GB	10.6 tok/s	Runs well	26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking.	GGUF (opens in a new tab)
`F16`	35.8 GB	72.0 GB	4.9 tok/s	Slow	26Basic quality (D) · 26 of 100: how capable the model is at F16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking.	Original weights (opens in a new tab)

Memory figures assume 4K of context. Longer conversations or documents grow the KV cache.

Other hardware that runs it

26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 3090 Q8_0 · 44.9 tok/s · Runs great
26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 4090 Q8_0 · 48.4 tok/s · Runs great
26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. AMD Radeon RX 7900 XTX Q8_0 · 39.7 tok/s · Runs great
24Basic quality (D) · 24 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. AMD Radeon RX 9070 XT Q4_K_M · 52.0 tok/s · Runs great
24Basic quality (D) · 24 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 4070 Ti SUPER Q4_K_M · 63.4 tok/s · Runs great
24Basic quality (D) · 24 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 4080 SUPER Q4_K_M · 69.4 tok/s · Runs great

Quick answers

How much memory does Llama 2 13B Chat need on the AMD Ryzen AI Max+ 395 96GB?

At Q3_K_M it needs about 16.1 GB. The AMD Ryzen AI Max+ 395 96GB has 72.0 GB usable.

How fast is Llama 2 13B Chat on the AMD Ryzen AI Max+ 395 96GB?

About 26.9 tok/s at Q3_K_M, an estimate from the hardware’s memory bandwidth and the size of the model.

Which quantization of Llama 2 13B Chat should I use on the AMD Ryzen AI Max+ 395 96GB?

Q3_K_M. Of the versions that fit, it is the one that runs best; when two run equally well, the check prefers the higher-rated one.

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.