AIbase
EN
Home
AI Models Finder
Llama 4 Scout
L

Llama 4 Scout

Llama 4 Scout is a native multimodal model capable of processing text and images. It features an expert (MoE) architecture with 17 billion active parameters (109B in total), consisting of 16 experts, and supports various multimodal tasks such as dialogue interaction, image analysis, and code generation. The model includes a context window of 10 million tokens.
Intelligence(Medium)
Speed(Medium)
Input Supported Modalities
Yes
Is Reasoning Model
10,000,000
Context Window
10,000,000
Maximum Output Tokens
2024-08-01
Knowledge Cutoff
Go Compare

Pricing

￥0.58 /M tokens
Input
￥2.16 /M tokens
Output
￥1.89 /M tokens
Blended Price

Quick Simple Comparison

Input

Output

Llama 4 Scout
￥0.08
Llama 4 Maverick
￥0.17
Llama 3.2 Instruct 1B

Basic Parameters

Llama 4 ScoutTechnical Parameters
Parameter Count
109,000.0M
Context Length
10.0M tokens
Training Data Cutoff
2024-08-01
Open Source Category
Open Weights (Permissive License)
Multimodal Support
Text, Image
Throughput
776
Release Date
2025-04-05
Response Speed
127.073,906 tokens/s

Benchmark Scores

Below is the performance of Llama 4 Scout in various standard benchmark tests. These tests evaluate the model's capabilities in different tasks and domains.
Intelligence Index
42.99
Large Language Model Intelligence Level
Coding Index
23.48
Indicator of AI model performance on coding tasks
Math Index
56.37
Capability indicator in solving mathematical problems, mathematical reasoning, or performing math-related tasks
MMLU Pro
75.2
Massive Multitask Multimodal Understanding - Testing understanding of text, images, audio, and video
GPQA
58.7
Graduate Physics Questions Assessment - Testing advanced physics knowledge with diamond science-level questions
HLE
4.3
The model's comprehensive average score on the Hugging Face Open LLM Leaderboard
LiveCodeBench
29.9
Specific evaluation focused on assessing large language models' ability in real-world code writing and solving programming competition problems
SciCode
17
The model's capability in code generation for scientific computing or specific scientific domains
HumanEval
82.6
Score achieved by the AI model on the specific HumanEval benchmark test set
Math 500 Score
84.4
Score on the first 500 larger, more well-known mathematical benchmark tests
AIME Score
28.3
An indicator measuring an AI model's ability to solve high-difficulty mathematical competition problems (specifically AIME level)
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase