MODEL / META

Llama 3.2 3B

Long advertised context still increases KV-cache memory; use modest context on small devices.

总参数3B

估算内存~2.8 GBGGUF Q4

上下文128KText

LicenseLlama 3.2 Community8 official languages

01 / 适合

02 / 估算适配

03 / 部署方式

OllamaEasyollama run llama3.2:3b

llama.cppMediumllama-server -m /path/to/verified-model.gguf --port 8080

LM StudioEasySearch “meta-llama/Llama-3.2-3B-Instruct” → Load a Q4/INT4 build → Start Server

MLX LMMediummlx_lm.chat --model meta-llama/Llama-3.2-3B-Instruct

注意

Long advertised context still increases KV-cache memory; use modest context on small devices.