MiniMax’s latest update introduces a Cover feature that transforms any song into a new style within 20 seconds, alongside the Music 2.6 model which boosts speed, expands instrument variety and enhances mid‑low‑frequency quality, with open‑source Music Skills and playlist agents enabling fully controllable AI‑generated music.
Xiaomi announced that a suite of AI research papers—including a large‑scale audio‑text dataset, a federated learning framework for domain and class generalization, a dual‑encoder music evaluation model, a cross‑domain audio‑text pre‑training system, a one‑step video‑to‑audio synthesis method, a training‑free frame‑selection technique for long‑video understanding, and a unified multimodal retrieval architecture—were accepted to the prestigious ICASSP 2026 conference, showcasing detailed methodologies, benchmark results, and potential impact across audio, vision, and multimodal AI applications.
The article reviews three open‑source AI tools—Claude Cowork for file‑based assistance, the LTX‑2 video generation model runnable on 8 GB GPUs via Pinokio, and SongGeneration Studio for end‑to‑end music creation—detailing their features, performance benchmarks, and usage steps for creators.
Meta introduced AudioCraft, a suite of generative AI models—including MusicGen, AudioGen, and EnCodec—that can create high‑quality music and sound effects from text prompts, offering open‑source weights and code for researchers and positioning the technology as a potential new standard instrument.
Meta’s chief AI scientist Yann LeCun warned that current large language models still fall short of human and even dog intelligence, citing their lack of real‑world understanding, while Meta unveiled three new generative AI models—Voicebox for speech, MusicGen for music, and I‑JEPA for image reasoning—showcasing both progress and remaining limitations.
musicpy is a Python library that lets developers write music using concise, theory‑based syntax, outputting MIDI files, and aims to enable AI composition through explicit music‑theory rules rather than deep‑learning, offering classes for notes, chords, scales and supporting mathematical operations on musical data.
