MODEL / META

Llama 4 Scout

Meta's single-H100 fit requires INT4. All 109B parameters must be stored; 17B active is not the memory number.

總參數109B17B active / token

估算記憶體~78 GBINT4 MoE

上下文10M advertisedText + Image

LicenseLlama 4 CommunityMultilingual

01 / 適合

02 / 估算適配

03 / 部署方式

vLLMAdvancedvllm serve meta-llama/Llama-4-Scout-17B-16E-Instruct --dtype auto

TensorRT-LLMAdvancedtrtllm-serve meta-llama/Llama-4-Scout-17B-16E-Instruct

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

注意

Meta's single-H100 fit requires INT4. All 109B parameters must be stored; 17B active is not the memory number.