Back to leaderboard

google/gemma-4-e4b

LM-STUDIO GGUF

gemma4 · Q4_K_M

Good

Jul 7, 2026 · AMD Ryzen 7 5800H

gemma-3-4b-it-q4-k-m

OLLAMA GGUF

llama · 4B · Q4_K_M

Excellent

Jul 9, 2026 · Intel Core™ i5-10400

Global Score
68 vs 80
Hardware Fit
93 vs 100
Quality Score
57 vs 71

Hardware

google/gemma-4-e4b gemma-3-4b-it-q4-k-m
MachineLENOVO 82JQMac Pro
CPUAMD Ryzen 7 5800HIntel Core™ i5-10400
Cores1612
RAM16 GB32 GB
GPUNVIDIA GeForce RTX 3060 Laptop GPUAMD Radeon RX 6700 XT
OSMicrosoft Windows 11 家庭版 中文版 10.0.26100macOS 26.5.1
Archx64x64
Power Modebalancedbalanced

Performance

google/gemma-4-e4b gemma-3-4b-it-q4-k-m
Tokens/sec26.275.0
First chunk17 ms136 ms
TTFT572 ms136 ms
Load timeN/A0.0 s
Memory usage6.5 GB0.8 GB
Memory %41%3%

HW Fit Score Breakdown

google/gemma-4-e4b

Speed
46/50
TTFT
20/20
Memory
27/30

gemma-3-4b-it-q4-k-m

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

Quality

google/gemma-4-e4b

Reasoning
17/20
Coding
19/20
Instruction
4/20
Structured
4/15
Math
12/15
Multilingual
1/10
Reasoning: Strong Coding: Strong Instruction Following: Poor Structured Output: Weak Math: Strong Multilingual: Poor

gemma-3-4b-it-q4-k-m

Reasoning
11/20
Coding
17/20
Instruction
14/20
Structured
14/15
Math
5/15
Multilingual
10/10
Reasoning: Adequate Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Weak Multilingual: Strong
View google/gemma-4-e4b details | View gemma-3-4b-it-q4-k-m 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