Back to leaderboard

deepseek-r1:7b

OLLAMA GGUF

qwen2 · 7.6B · Q4_K_M

Good

Mar 5, 2026 · Intel Core™ i5-6500

qwen3.5-9b-mlx

LM-STUDIO MLX

qwen3_5 · 9B · 4bit

Good

Mar 12, 2026 · Apple M4 Pro

Global Score
66 vs 77
Hardware Fit
63 vs 67
Quality Score
67 vs 81

Hardware

deepseek-r1:7b qwen3.5-9b-mlx
MachineGigabyte Technology Co., Ltd. H170M-D3HMac mini
CPUIntel Core™ i5-6500Apple M4 Pro
Cores414
RAM16 GB64 GB
GPUGP104 [GeForce GTX 1070]Apple M4 Pro
OSUbuntu 24.04.4 LTSmacOS 15.7.4
Archx64arm64
Power Modelow-powerbalanced

Performance

deepseek-r1:7b qwen3.5-9b-mlx
Tokens/sec35.620.3
First chunk224 ms4216 ms
TTFT7.5 s4.2 s
Load time7.7 sN/A
Memory usage4.6 GB5.6 GB
Memory %29%9%

HW Fit Score Breakdown

deepseek-r1:7b

Speed
50/50
TTFT
7/20
Memory
6/30

qwen3.5-9b-mlx

Speed
29/50
TTFT
8/20
Memory
30/30

Quality

deepseek-r1:7b

Reasoning
16/20
Coding
10/20
Instruction
9/20
Structured
10/15
Math
14/15
Multilingual
8/10
Reasoning: Strong Coding: Adequate Instruction Following: Weak Structured Output: Adequate Math: Strong Multilingual: Strong

qwen3.5-9b-mlx

Reasoning
15/20
Coding
18/20
Instruction
16/20
Structured
14/15
Math
8/15
Multilingual
10/10
Reasoning: Adequate Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Adequate 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