Meta Llama 2 Release Fuels AI Surge as Apple, New Global Models Emerge

Meta Platforms released Llama 2, an open‑weight large language model that permits commercial use, prompting rapid adoption and a wave of derivative models that quickly topped leaderboards. The company’s decision follows earlier limited releases of LLaMA 1 and aims to pre‑empt weight leaks while positioning Meta competitively.

Bloomberg reported that Apple is quietly developing its own GPT‑style chatbot, dubbed “Ajax,” indicating the tech giant’s entry into the generative‑AI race. At the same time, several new AI models were announced worldwide: the 9.75‑billion‑parameter multimodal model Inkling from the former OpenAI CTO’s Thinking Machines Lab; Laguna S 2.1, a coding‑focused 118‑billion‑parameter mixture‑of‑experts model that runs on a single NVIDIA DGX Spark; Alibaba’s ABot‑World‑0, a world‑generation model that can run in real‑time on a desktop RTX 5090; and Qwen‑3 6‑27B‑Fable‑Fusion‑711, an uncensored model fine‑tuned for higher benchmark scores.