Llama 4 Maverick
Meta released the 多模态 model · Launched on 2025-04
Open Source
多模态
开源
API 可用
Llama 4 Maverick is Meta's high-performance open-source multimodal model, also using MoE architecture, with 400B total parameters and 17B active parameters. It excels in image understanding, coding, and reasoning, with performance comparable to GPT-4o. Maverick has an ultra-long context window of 10 million tokens, providing significant advantages in long-document analysis, codebase understanding, and multimodal dialogue.
核心特性
- 1000万 tokens 超长上下文
- 业界领先的多模态
- MoE 高效推理
- 开源可商用
核心优势
- 多模态能力最强
- 上下文窗口极大
- 推理高效
参数量
400B (17B 激活)
上下文窗口
10M tokens
API 支持
Yes
开源状态
Open Source
发布日期
2025-04
提供商
Meta
免费开源
MMLU 87.2%, LMArena 综合排名与 GPT-4o 相当
社区数据
Likes
514
Downloads
9,661
License
other
Task
image-text-to-text
Repo updated
2025-05-22
huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct
