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

qwen3:14b

OLLAMA GGUF

qwen3 · 14.8B · Q4_K_M

Excellent

Jun 13, 2026 · AMD Ryzen 9 3900X 12-Core Processor

Global Score
51 vs 82
Hardware Fit
31 vs 90
Quality Score
60 vs 79

Hardware

llama3.3:70b-instr… qwen3:14b
MachineDell Inc. Precision Tower 7910Micro-Star International Co., Ltd. MS-7C37
CPUIntel Xeon® E5-2637 v3AMD Ryzen 9 3900X 12-Core Processor
Cores1624
RAM96 GB64 GB
GPUNVIDIA GeForce RTX 3090, NVIDIA GeForce RTX 3060GameViewer Virtual Display Adapter, NVIDIA GeForce RTX 3080
OSMicrosoft Windows 10 Pro 10.0.19045Microsoft Windows 11 专业版 10.0.26200
Archx64x64
Power Modebalancedbalanced

Performance

llama3.3:70b-instr… qwen3:14b
Tokens/sec1.774.4
First chunk5553 ms324 ms
TTFT5.6 s2.6 s
Load time0.4 s5.1 s
Memory usage50.3 GB9.0 GB
Memory %53%14%

HW Fit Score Breakdown

llama3.3:70b-instruct-q…

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

qwen3:14b

Speed
50/50
TTFT
10/20
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

qwen3:14b

Reasoning
18/20
Coding
11/20
Instruction
17/20
Structured
14/15
Math
10/15
Multilingual
9/10
Reasoning: Strong Coding: Adequate 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