Back to leaderboard

llama3.2:3b

OLLAMA GGUF

llama · 3.2B · Q4_K_M

Good

Jun 25, 2026 · Cortex-A78AE

llama3.3:70b

OLLAMA GGUF

llama · 70.6B · Q4_K_M

Not Rec.

Jun 26, 2026 · Cortex-X925

Global Score
74 vs 73
Hardware Fit
99 vs 51
Quality Score
63 vs 83

Hardware

llama3.2:3b llama3.3:70b
MachineNVIDIA NVIDIA Jetson Orin Nano Engineering Reference Developer Kit SuperNVIDIA NVIDIA_DGX_Spark
CPUCortex-A78AECortex-X925
Cores620
RAM7 GB122 GB
GPUIntegrated / UnknownDevice 2e12
OSUbuntu 22.04.5 LTSUbuntu 24.04.4 LTS
Archarm64arm64
Power Modebalancedperformance

Performance

llama3.2:3b llama3.3:70b
Tokens/sec21.14.6
First chunk1032 ms560 ms
TTFT1.0 s560 ms
Load time8.4 s13.8 s
Memory usage2.4 GB49.5 GB
Memory %32%41%

HW Fit Score Breakdown

llama3.2:3b

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

llama3.3:70b

Speed
4/50
TTFT
20/20
Memory
27/30

Quality

llama3.2:3b

Reasoning
10/20
Coding
14/20
Instruction
15/20
Structured
12/15
Math
3/15
Multilingual
9/10
Reasoning: Adequate Coding: Adequate Instruction Following: Strong Structured Output: Strong Math: Poor Multilingual: Strong

llama3.3:70b

Reasoning
17/20
Coding
18/20
Instruction
12/20
Structured
15/15
Math
11/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Adequate 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