Meta has launched the primary two fashions from its multimodal Llama 4 suite: LLama 4 Scout and Llama 4 Maverick. Maverick is “the workhorse” of the 2 and excels at picture and textual content understanding for “normal assistant and chat use instances,” the corporate stated in a weblog put up, whereas the smaller mannequin Scout might deal with issues like “multi-document summarization, parsing intensive consumer exercise for personalised duties, and reasoning over huge codebases.” The corporate additionally launched Llama 4 Behemoth, an upcoming mannequin it says is “among the many world’s smartest LLMs” — and CEO Mark Zuckerberg stated we’ll be listening to a couple of fourth mannequin, LLama 4 Reasoning, “within the subsequent month.”
Each Maverick and Scout can be found to obtain now from the LLama web site and Hugging Face, they usually’ve been added to Meta AI, together with for WhatsApp, Messenger and Instagram DMs.
Scout has 17 billion energetic parameters with 16 specialists, Meta says. In line with Zuckerberg, “It’s extraordinarily quick, natively multimodal, and has an trade main, almost infinite 10 million token context size, and it’s designed to run on a single GPU.” Maverick then again has 17 billion energetic parameters with 128 specialists. The corporate says it beats rivals like GPT-4o and Gemini 2.0 on coding, reasoning, multilingual, long-context and picture benchmarks, and stacks up in opposition to DeepSeek v3.1 on reasoning and coding.
Zuckerberg is already calling the upcoming Behemoth mannequin, which continues to be coaching, “the best performing base mannequin on the earth,” with 288 billion energetic parameters, in keeping with the corporate. It will not be right here but, nevertheless it’s probably we’ll be listening to much more about that and the Reasoning mannequin quickly; Meta’s large AI developer convention, LlamaCon, is just some weeks away.