TAAFT
Free mode
100% free
Freemium
Free Trial
Create tool
ModelsLlama 4 Behemoth

Llama 4 Behemoth

By Meta
Multimodal Gen
Released: April 5, 2025

Overview

Description

Large Multimodal Model (Teacher); MoE architecture (16 experts, 288B active, ~2T total params); Used to distill Llama 4 Scout/Maverick; Outperforms GPT-4.5/Claude 3.7 Sonnet/Gemini 2.0 Pro on STEM benchmarks; Still in training (as of Apr 2025).

About Meta

We're connecting people to what they care about, powering new, meaningful experiences, and advancing the state-of-the-art through open research and accessible tooling.

Location: California, US
Website: ai.meta.com
View Company Profile

Related Models

Llama 3.1 (405B)

No description available.

Text

Mistral Small 3.2

Mistral Small 3.2 (24B) is the largest configuration of the “Small” line—around 24B dense parameters—offering stronger reasoning and coding skills while staying cheaper and faster than Mistral Medium or Large. It supports long context, JSON outputs, and tool/function calling for production copilots and RAG.

Multimodal New

Llama Guard 3

Llama Guard 3 is Meta’s open-weight safety model for moderating AI prompts and responses. It classifies policy risks, returns structured JSON (allow/block with categories and rationale), and is designed to run beside your assistant for low-latency, context-aware guardrails.

Text New
Last updated: April 15, 2025