Creator

Meta AI (FAIR)

Released

2025-04-05

Intelligence

10.0

Artificial Analysis Index

Coding

8.2

Artificial Analysis Index

In $/1M

$0.18

input tokens

Out $/1M

$0.66

output tokens

Blended $/1M

$0.30

3:1 blended

Speed

82

tokens / sec

Profile

License, openness, and modality — the governance layer.

License

Llama 4 Community License Agreement

non-commercial

Weights

Open

downloadable

Modalities

Image · Text

Parameters

109B

total

Llama 4 Scout is a natively multimodal model capable of processing both text and images. It features a 17 billion activated parameter (109B total) mixture-of-experts (MoE) architecture with 16 experts, supporting a wide range of multimodal tasks such as conversational interaction, image analysis, and code generation. The model includes a 10 million token context window.

Capability profile

Category strength across 20 domains, via LLM Stats.

image to text
90legal
80vision
80finance
80language
80multimodal
80math
70
general
70reasoning
70healthcare
70biology
60physics
60chemistry
60code
30
Via LLM Stats

Benchmark breakdown

Independent evaluation scores, normalized to 0–100.

GPQA Diamond
59
Humanity’s Last Exam
4
MMLU-Pro
75
SciCode
17
LiveCodeBench
30
MATH-500
84
AIME 2025
14
τ²-Bench (agentic)
15
Terminal-Bench Hard
2
IFBench
40
Via Artificial Analysis
© 2026 NYSGPT2525 LLC