What is Llama 4 Scout?

Llama 4 Scout is cataloged under Meta / Llama with canonical ID meta/llama-4-scout.

MEMeta/ Llama

Llama 4 Scout

meta/llama-4-scout

Llama 4 Scout 17B Instruct (16E) is a mixture-of-experts (MoE) language model developed by Meta, activating 17 billion parameters out of a total of 109B. It supports native multimodal input...

Compare

Model information

Released

2025-04-05

Context window

1.3M

Max output

16K

Model parameters

109B

Weights

Open

Modalities

Input

ImageText

Output

Text

Identity

Version: llama-4-scout
Knowledge cutoff: 2024-08-31
License: llama4

Resources

Hugging Face

Capabilities

Open weightStructuredTemperatureTool call

Providers

Provider	Provider model ID	Protocol	Context	Max output	Capabilities	Notes	Pricing	Streaming
OpenRouter	meta-llama/llama-4-scout	openai_chat	1.3M	16K	AttachmentsTool callStructuredTemperature	Hugging Face: meta-llama/Llama-4-Scout-17B-16E-Instruct	input $0.1/M · output $0.3/M	Yes
NanoGPT	meta-llama/llama-4-scout	openai_chat	1.3M	16K	AttachmentsTool callStructured	-	input $0.085/M · output $0.46/M cache read $0.043/M	Yes
Kilo Gateway	meta-llama/llama-4-scout	openai_chat	1.3M	16K	AttachmentsTool callStructuredTemperature	-	input $0.1/M · output $0.3/M	Yes
Meta	llama-4-scout	custom	1.3M	16K	-	-	-	Yes