Back to leaderboard

llama3.3:70b-instruct-q4_K_M

OLLAMA GGUF

llama · 70.6B · Q4_K_M

Not Rec.

Aug 20, 2026 · Intel Xeon® E5-2637 v3

qwen3.5-2b-mlx

LM-STUDIO MLX

qwen3_5 · 2B · 4bit

Good

Mar 6, 2026 · Apple M4 Pro

Global Score
51 vs 65
Hardware Fit
31 vs 88
Quality Score
60 vs 55

Hardware

llama3.3:70b-instr… qwen3.5-2b-mlx
MachineDell Inc. Precision Tower 7910Mac mini
CPUIntel Xeon® E5-2637 v3Apple M4 Pro
Cores1614
RAM96 GB64 GB
GPUNVIDIA GeForce RTX 3090, NVIDIA GeForce RTX 3060Apple M4 Pro
OSMicrosoft Windows 10 Pro 10.0.19045macOS 15.7.4
Archx64arm64
Power Modebalancedbalanced

Performance

llama3.3:70b-instr… qwen3.5-2b-mlx
Tokens/sec1.748.9
First chunk5553 ms4208 ms
TTFT5.6 s4.2 s
Load time0.4 sN/A
Memory usage50.3 GB1.6 GB
Memory %53%3%

HW Fit Score Breakdown

llama3.3:70b-instruct-q…

Speed
2/50
TTFT
5/20
Memory
24/30

qwen3.5-2b-mlx

Speed
50/50
TTFT
8/20
Memory
30/30

Quality

llama3.3:70b-instruct-q…

Reasoning
17/20
Coding
6/20
Instruction
6/20
Structured
11/15
Math
11/15
Multilingual
9/10
Reasoning: Strong Coding: Weak Instruction Following: Weak Structured Output: Adequate Math: Adequate Multilingual: Strong

qwen3.5-2b-mlx

Reasoning
10/20
Coding
7/20
Instruction
13/20
Structured
14/15
Math
3/15
Multilingual
8/10
Reasoning: Adequate Coding: Weak Instruction Following: Adequate Structured Output: Strong Math: Poor Multilingual: Strong

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