hf.co/unsloth/Qwen3.8-27B-GGUF:UD-IQ1_M

OLLAMA GGUF

qwen35 · 27.3B · unknown

Good

Aug 21, 2026 · AMD RYZEN AI MAX+ 395

mlx-community/meta-llama-3.1-8b-instruct

LM-STUDIO MLX

llama · 8B · 4bit

Good

Mar 5, 2026 · Apple M4 Pro

Global Score

68 vs 77

Hardware Fit

69 vs 100

Quality Score

67 vs 67

Hardware

hf.co/unsloth/Qwen… mlx-community/meta…

MachineAZW GTR ProMac mini

CPUAMD RYZEN AI MAX+ 395Apple M4 Pro

Cores3214

RAM125 GB64 GB

GPURadeon 8060SApple M4 Pro

OSUbuntu 24.04.4 LTSmacOS 15.7.4

Archx64arm64

Power Modelow-powerbalanced

Performance

hf.co/unsloth/Qwen… mlx-community/meta…

Tokens/sec21.154.3

First chunk571 ms277 ms

TTFT571 ms277 ms

Load time4.9 sN/A

Memory usage22.7 GB4.2 GB

Memory %18%7%

HW Fit Score Breakdown

hf.co/unsloth/Qwen3.8-2…

Speed

20/50

TTFT

19/20

Memory

30/30

mlx-community/meta-llam…

Speed

50/50

TTFT

20/20

Memory

30/30

Quality

hf.co/unsloth/Qwen3.8-2…

Reasoning

12/20

Coding

10/20

Instruction

10/20

Structured

15/15

Math

10/15

Multilingual

10/10

Reasoning: Adequate Coding: Weak Instruction Following: Adequate Structured Output: Strong Math: Adequate Multilingual: Strong

mlx-community/meta-llam…

Reasoning

11/20

Coding

14/20

Instruction

14/20

Structured

15/15

Math

4/15

Multilingual

9/10

Reasoning: Adequate Coding: Adequate Instruction Following: Adequate Structured Output: Strong Math: Poor Multilingual: Strong

View hf.co/unsloth/Qwen… details | View mlx-community/meta… details

Run yours and compare

$ npm install -g metrillm@latest

$ metrillm

Requires Node 20+ and Ollama or LM Studio running

Or run without installing: npx metrillm@latest