Llama 4 Scout
Class-leading natively multimodal model that offers superior text and visual intelligence, single H100 GPU efficiency, and a 10M context window for seamless long document analysis.
Llama 4 Scout is a specific Llama 4 model release from Meta: a natively multimodal mixture-of-experts model with 17B active parameters, 16 experts, 109B total parameters, multilingual text-and-image input, text/code output, and a 10M-token context window.
Pricing
Model profile · Current snapshot
Input / 1M
$0.18
Output / 1M
$0.66
Blended / 1M
$0.30
Output TPS
98.33
TTFT (s)
0.61
Model Intelligence
Context window
10,000,000 tokens
Arena ranking
10
Benchmarkable
Yes
Model level
release
Intelligence Index
10.3
Coding Index
8.2
Math Index
14
MMLU Pro
0.75
GPQA
0.59
HLE
0.04
LiveCodeBench
0.3
SciCode
0.17
MATH-500
0.84
AIME
0.28
AIME 2025
0.14
IFBench
0.4
LCR
0.3
TerminalBench Hard
0.02
TAU2
0.15
Recent stories
0 linked stories
No linked stories yet.