Back to leaderboard

qwen3:4b

OLLAMA GGUF

qwen3 · 4.0B · Q4_K_M

Not Rec.

Jun 26, 2026 · Cortex-A78AE

Kimi-48B:48B

OLLAMA GGUF

kimi-linear · 49.1B · unknown

Excellent

Aug 22, 2026 · NVIDIA RTX Spark N1X (6144-core GPU, 20-core ) RTX Spark N1X (6144-core GPU, 20-core )

Global Score
29 vs 90
Hardware Fit
86 vs 96
Quality Score
4 vs 88

Hardware

qwen3:4b Kimi-48B:48B
MachineNVIDIA NVIDIA Jetson Orin Nano Engineering Reference Developer Kit SuperHP HyperX OMEN Transcend Gaming Laptop 14-ga0xxx5CD612DXML
CPUCortex-A78AENVIDIA RTX Spark N1X (6144-core GPU, 20-core ) RTX Spark N1X (6144-core GPU, 20-core )
Cores620
RAM7 GB55 GB
GPUIntegrated / UnknownNVIDIA RTX Spark N1X (6144-core Blackwell RTX GPU), NVIDIA NPU
OSUbuntu 22.04.5 LTSMicrosoft Windows 11 Pro 10.0.28000
Archarm64arm64
Power Modebalancedbalanced

Performance

qwen3:4b Kimi-48B:48B
Tokens/sec17.962.7
First chunk1010 ms397 ms
TTFT6.4 s397 ms
Load time9.5 s16.8 s
Memory usage3.0 GB25.5 GB
Memory %40%47%

HW Fit Score Breakdown

qwen3:4b

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

Kimi-48B:48B

Speed
50/50
TTFT
20/20
Memory
26/30

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

Kimi-48B:48B

Reasoning
16/20
Coding
19/20
Instruction
16/20
Structured
15/15
Math
12/15
Multilingual
10/10
Reasoning: Strong Coding: Strong Instruction Following: Strong Structured Output: Strong Math: Strong Multilingual: Strong
View qwen3:4b details | View Kimi-48B:48B 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