Back to leaderboard

qwen3:4b

OLLAMA GGUF

qwen3 · 4.0B · Q4_K_M

Not Rec.

Jun 26, 2026 · Cortex-A78AE

qwen3-next:80b-a3b-instruct-q4_K_M

OLLAMA GGUF

qwen3next · 79.7B · Q4_K_M

Not Rec.

Aug 23, 2026 · Intel Xeon® E5-4650 v2

Global Score
29 vs 74
Hardware Fit
86 vs 42
Quality Score
4 vs 87

Hardware

qwen3:4b qwen3-next:80b-a3b…
MachineNVIDIA NVIDIA Jetson Orin Nano Engineering Reference Developer Kit SuperDell Inc. PowerEdge R820
CPUCortex-A78AEIntel Xeon® E5-4650 v2
Cores680
RAM7 GB256 GB
GPUIntegrated / UnknownMicrosoft Basic Display Adapter
OSUbuntu 22.04.5 LTSMicrosoft Windows Server 2022 Datacenter 10.0.20348
Archarm64x64
Power Modebalancedbalanced

Performance

qwen3:4b qwen3-next:80b-a3b…
Tokens/sec17.93.9
First chunk1010 ms2089 ms
TTFT6.4 s2.1 s
Load time9.5 s73.3 s
Memory usage3.0 GB46.9 GB
Memory %40%18%

HW Fit Score Breakdown

qwen3:4b

Speed
50/50
TTFT
8/20
Memory
28/30

qwen3-next:80b-a3b-inst…

Speed
3/50
TTFT
9/20
Memory
30/30

Quality

qwen3:4b

Reasoning
4/20
Coding
0/20
Instruction
0/20
Structured
0/15
Math
0/15
Multilingual
0/10
Reasoning: Poor Coding: Poor Instruction Following: Poor Structured Output: Poor Math: Poor Multilingual: Poor

qwen3-next:80b-a3b-inst…

Reasoning
18/20
Coding
19/20
Instruction
12/20
Structured
15/15
Math
13/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Strong 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