Skip to main content
ModelLLMsvLlama-4-Scout-17B-16E-Instruct

Llama 4 Scout

by Meta · open-source · Last verified 2026-03-17

Meta's fourth-generation mixture-of-experts model with 17B active parameters from 109B total, featuring native multimodal understanding and a 10M token context window. The most efficient Llama model to date.

https://llama.meta.com
B
BAbove Average
Adoption: B+Quality: AFreshness: A+Citations: B+Engagement: F

Specifications

License
Llama 4 Community License
Pricing
open-source
Capabilities
text-generation, code-generation, multimodal-vision, tool-use, long-context, reasoning
Integrations
huggingface, ollama, vllm, together-ai, aws-bedrock, groq
Use Cases
long-document-analysis, multimodal-reasoning, code-generation, agentic-workflows, enterprise-deployment
API Available
Yes
Parameters
109B (17B active)
Context Window
10M tokens
Modalities
text, image, video
Training Cutoff
August 2025
Tags
llm, open-source, moe, multimodal, long-context, meta
Added
2026-03-17
Completeness
100%

Index Score

66
Adoption
76
Quality
88
Freshness
95
Citations
72
Engagement
0

Explore the full AI ecosystem on Agents as a Service