james-see/ltx-video-mac

Native macOS app for local AI video on Apple Silicon — LTX-2, LTX-2.3, LTX-2.5, and MiniMax H3

What it solves

This project provides a native macOS application that allows users to generate AI videos locally on Apple Silicon Macs. It eliminates the need for cloud-based video generation by bringing high-performance models like LTX-2, LTX-2.5, and MiniMax H3 directly to the Mac, while integrating audio generation and voiceover tools into a single interface.

How it works

The app is built with SwiftUI and uses Apple's MLX framework to run video models efficiently on M-series chips. It acts as a bridge to various Python-based backends (such as mlx-video-with-audio and h3.c), managing model downloads from Hugging Face and handling the generation queue. It also includes an optional Gemma-powered prompt enhancer to rewrite simple prompts into detailed, optimized descriptions for better visual and audio results.

Who it’s for

Content creators and AI enthusiasts with Apple Silicon Macs (M1 through M4) who want to generate high-quality video clips locally with integrated audio, voiceover, and background music.

Highlights

  • Multi-Model Support: Supports LTX-2, LTX-2.3, LTX-2.5, and MiniMax H3.
  • Text-to-Video & Image-to-Video: Generates clips from text prompts or animates existing images using keyframes.
  • Integrated Audio: Built-in synchronized audio generation, local MLX-Audio voiceovers, and ElevenLabs cloud integration for music and narration.
  • Local REST API: Includes a server for queuing generations via external requests.
  • Mac Native Experience: Features a SwiftUI interface with a generation queue, history management, and parameter presets.
  • Prompt Enhancement: Optional Gemma-driven rewriting to optimize prompts for the LTX-2 series.

Related

  • Project
    jjang-ai/mlxstudio

    A native macOS app for Apple Silicon that provides a user-friendly interface to run LLMs, vision models, and image generation locally using the MLX framework.

  • Project
    Blaizzy/mlx-audio

    An audio processing library optimized for Apple Silicon that provides fast inference for text-to-speech, speech-to-text, speech-to-speech, and music generation.

  • Project
    raullenchai/Rapid-MLX

    A high-performance LLM inference server for Apple Silicon that optimizes MLX for faster local inference and reliable tool-calling for coding agents.

  • Project
    Lightricks/LTX-2

    LTX‑2 is an open‑source diffusion‑based audio‑video foundation model (DiT architecture) that generates synchronized video and sound from text prompts. It provides fast distilled and high‑fidelity pipelines, supports 4K HDR, and ships with modular components (transformer, Gemma‑4 text encoder, video/audio VAEs, up‑scalers, LoRA adapters). The repo includes CLI tools, Python APIs, and integration with ComfyUI, making it suitable for both research prototyping and production‑level video generation.

  • Project
    Blaizzy/mlx-audio-swift

    A modular Swift SDK for audio processing on Apple Silicon, providing a unified API for TTS, STT, VAD, and audio codecs using the MLX framework.