qwen3.5-2b-mlx

LM-STUDIO MLX

qwen3_5 · 2B · 4bit

Good

Mar 6, 2026 · Apple M4 Pro

deepseek-r1:1.5b

OLLAMA GGUF

qwen2 · 1.8B · Q4_K_M

Not Rec.

Apr 29, 2026 · Intel Core™ i5-8350U

Global Score

65 vs 37

Hardware Fit

88 vs 73

Quality Score

55 vs 22

Hardware

qwen3.5-2b-mlx deepseek-r1:1.5b

MachineMac miniLENOVO 20L7CTO1WW

CPUApple M4 ProIntel Core™ i5-8350U

Cores148

RAM64 GB23 GB

GPUApple M4 ProKaby Lake-R GT2 [UHD Graphics 620]

OSmacOS 15.7.4Fedora Linux 43

Archarm64x64

Power Modebalancedlow-power

Performance

qwen3.5-2b-mlx deepseek-r1:1.5b

Tokens/sec48.918.4

First chunk4208 ms593 ms

TTFT4.2 s30.0 s

Load timeN/A3.8 s

Memory usage1.6 GB1.0 GB

Memory %3%4%

HW Fit Score Breakdown

qwen3.5-2b-mlx

Speed

50/50

TTFT

8/20

Memory

30/30

deepseek-r1:1.5b

Speed

43/50

TTFT

0/20

Memory

30/30

Quality

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

deepseek-r1:1.5b

Reasoning

4/20

Coding

0/20

Instruction

5/20

Structured

3/15

Math

7/15

Multilingual

3/10

Reasoning: Poor Coding: Poor Instruction Following: Weak Structured Output: Poor Math: Weak Multilingual: Weak

View qwen3.5-2b-mlx details | View deepseek-r1:1.5b 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