deepseek-v2:16b

OLLAMA GGUF

deepseek2 · 15.7B · Q4_0

Good

Mar 2, 2026 · Apple M4

falcon3-1b-instruct

LM-STUDIO MLX

llama · 1B · 3bit

Marginal

Mar 4, 2026 · Apple M4

Global Score

69 vs 54

Hardware Fit

100 vs 100

Quality Score

55 vs 34

Hardware

deepseek-v2:16b falcon3-1b-instruct

MachineMacBook AirMacBook Air

CPUApple M4Apple M4

Cores1010

RAM32 GB32 GB

GPUApple M4Apple M4

OSmacOS 26.3macOS 26.3

Archarm64arm64

Power Modebalancedbalanced

Performance

deepseek-v2:16b falcon3-1b-instruct

Tokens/sec56.6121.8

First chunkN/A176 ms

TTFT239 ms176 ms

Load time1.2 sN/A

Memory usage17.7 GB0.7 GB

Memory %55%2%

HW Fit Score Breakdown

deepseek-v2:16b

Speed

40/50

TTFT

30/20

Memory

30/30

falcon3-1b-instruct

Speed

50/50

TTFT

20/20

Memory

30/30

Quality

deepseek-v2:16b

Reasoning

7/20

Coding

9/20

Instruction

13/20

Structured

10/15

Math

8/15

Multilingual

8/10

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

falcon3-1b-instruct

Reasoning

7/20

Coding

1/20

Instruction

12/20

Structured

7/15

Math

2/15

Multilingual

5/10

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

View deepseek-v2:16b details | View falcon3-1b-instruct 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