ailexleon/Cydonia-24B-v4.3-mlx-fp16

TEXT GENERATIONConcurrency Cost:2Model Size:24BQuant:FP8Ctx Length:32kPublished:Dec 25, 2025Architecture:Transformer Cold

The ailexleon/Cydonia-24B-v4.3-mlx-fp16 is a 24 billion parameter language model, converted to the MLX format by ailexleon from TheDrummer/Cydonia-24B-v4.3. This model is designed for efficient inference on Apple silicon, leveraging the MLX framework. It maintains the capabilities of the original Cydonia-24B-v4.3 model, offering a substantial parameter count for general language generation tasks. Its primary utility lies in providing a readily deployable version for MLX-compatible environments.

Loading preview...

Model Overview

The ailexleon/Cydonia-24B-v4.3-mlx-fp16 is a 24 billion parameter language model, specifically converted for use with Apple's MLX framework. This model is a direct conversion of the TheDrummer/Cydonia-24B-v4.3 model, ensuring compatibility and optimized performance on Apple silicon.

Key Features

  • MLX Compatibility: Optimized for efficient inference on devices with Apple silicon, utilizing the mlx-lm library.
  • Parameter Count: Features 24 billion parameters, providing a robust foundation for complex language understanding and generation tasks.
  • Direct Conversion: Faithfully represents the capabilities of the original TheDrummer/Cydonia-24B-v4.3 model.

Usage

This model is intended for developers and researchers looking to leverage the Cydonia-24B-v4.3 model within the MLX ecosystem. It can be easily loaded and used for text generation with the mlx_lm library, as demonstrated in the provided Python code snippet. The conversion was performed using mlx-lm version 0.28.3.

Good For

  • Local inference on Apple silicon hardware.
  • Experimentation with large language models in an MLX environment.
  • Applications requiring a 24B parameter model with optimized performance on macOS.