Latest Content
Llama 4 Scout 17B 16E Instruct icon

Llama 4 Scout 17B 16E Instruct

NVIDIA
Llama-4-Scout-17B-16E-Instruct is an instruction-tuned, natively multimodal auto-regressive model developed by Meta, built on a mixture-of-experts architecture with 17B activated parameters (109B total). Designed for assistant-style and vision-language workloads, it supports multilingual text and image inputs and produces multilingual text and code outputs. The model features a native 10 million token context window, enabling extremely long-context analysis across documents, conversations, and multimodal inputs. It was trained on a mixture of publicly available and licensed data, including data from Meta’s products, with a knowledge cutoff of August 2024, and officially released on April 5, 2025.
TypeVision-Language Model
CapabilitiesText Generation, Instruction Following, Reasoning, Mathematical Reasoning+6 more
Links
Blog|HF Model Card
LicenseLlama4

Inference Instructions

Deploy and run this model on NVIDIA B200 GPUs using the command below. Copy the command to get started with inference.

CONSOLE
docker run -it --rm 
 --runtime=nvidia 
 --gpus all 
 --ipc=host 
 --shm-size=128g 
 -p 8000:8000 
 -v ~/.cache/huggingface:/root/.cache/huggingface 
 -e HF_TOKEN='YOUR_HF_TOKEN' 
 -e LD_LIBRARY_PATH='/usr/local/nvidia/lib64:/usr/local/nvidia/lib:/usr/lib/x86_64-linux-gnu' 
 vllm/vllm-openai:v0.15.0-cu130 
 meta-llama/Llama-4-Scout-17B-16E-Instruct 
  --tensor-parallel-size 8 
 --max-model-len auto 
  --max-num-batched-tokens 65536 
  --gpu-memory-utilization 0.95 
  --max-num-seqs 1024 
 --disable-log-requests 
 --trust-remote-code

Model Benchmarks

Each model was tested with a fixed input size and total token volume while increasing concurrency to measure serving performance under load.

ITL vs Concurrency

Time to First Token

Throughput Scaling

Total Tokens/sec vs Avg TTFT

Vultr Cloud GPU

NVIDIA HGX B200

Deploy NVIDIA B200 on Vultr Cloud GPU

Deploy NVIDIA B200 instancesContact Sales