meta-llama-3.1-8b-instruct

llama · 8B · 4bit

Good

Mar 3, 2026 · Apple M4

qwen3:4b

qwen3 · 4.0B · Q4_K_M

Marginal

Mar 1, 2026 · Apple M4

Global Score

76 vs 47

Hardware Fit

95 vs 53

Quality Score

68 vs 45

Hardware

meta-llama-3.1-8b-… qwen3:4b

MachineMacBook Air—

CPUApple M4Apple M4

Cores1010

RAM32 GB32 GB

GPUApple M4Apple M4

OSmacOS 26.3macOS 26.3

Archarm64arm64

Power Modebalancedunknown

Performance

meta-llama-3.1-8b-… qwen3:4b

Tokens/sec21.611.3

TTFT512 ms9.2 s

Load time0.0 s0.1 s

Memory usage0.0 GB7.2 GB

Memory %0%23%

HW Fit Score Breakdown

meta-llama-3.1-8b-instr…

Speed

35/40

TTFT

30/30

Memory

30/30

qwen3:4b

Speed

17/40

TTFT

6/30

Memory

30/30

Quality

meta-llama-3.1-8b-instr…

Reasoning

11/20

Coding

15/20

Instruction

14/20

Structured

14/15

Math

4/15

Multilingual

10/10

Reasoning: Adequate Coding: Strong Instruction Following: Adequate Structured Output: Strong Math: Poor Multilingual: Strong

qwen3:4b

Reasoning

4/20

Coding

6/20

Instruction

8/20

Structured

6/15

Math

12/15

Multilingual

9/10

Reasoning: Poor Coding: Weak Instruction Following: Weak Structured Output: Weak Math: Strong Multilingual: Strong

View meta-llama-3.1-8b-… details | View qwen3:4b details

Run yours and compare

npx metrillm@latest bench

Requires Node 20+ and Ollama or LM Studio running