infery
← All models

Llama 4 Scout

llama-4-scout

Chat & Textby Meta

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. 10,000,000 token context window, maximum output of 16,384 tokens. Higher uptime with 4 providers. Includes independent benchmarks from Artificial Analysis.

chatstreamingvisiontoolsjson mode

Details

Accepts
text
Context window
328K tokens

Pricing

Input
12.5 cr / 1M tokens
Output
37.5 cr / 1M tokens

Prices in credits (1 credit = $0.01).

Data schema

approximate

Input

FieldTypeDescription
promptrequiredstringThe text prompt.

Output

FieldTypeDescription
textstringGenerated text (choices[0].message.content).