Open Source Model Open Source

Llama

Llama is a family of open-weight, multimodal large language models from Meta, offering advanced capabilities for diverse AI applications.

9/10 MegaOne Score

Visit Llama →

Price Free / Open Source

Category Open Source Model

Model Open Source

Company Meta Platforms, Inc.

Team Size 75,472

Total Funding $2.3B

About Llama

Llama 4, released in April 2025, represents Meta's most advanced suite of open-weight, natively multimodal models, including Scout and Maverick. These models leverage a Mixture-of-Experts (MoE) architecture, providing enhanced efficiency and performance for tasks like image and text understanding, multilingual processing, coding, and agentic systems. Llama 4 Maverick is considered the best overall Llama model for general-purpose use, while Scout specializes in ultra-long context tasks with a 10 million token context window.

Why We Scored It 9/10

9/10

Llama maintains its position as a market-defining open-source model family with the release of Llama 4 variants and massive backing from Meta's resources ($2.3B funding, 70K employees). Despite some strategic uncertainty around open-source commitment, it remains a foundational model that has democratized AI access and continues to drive industry innovation.

Features & Platform

Free Tier

Llama Coverage