pa5haw/Mistral-Small-3.2-24B-Instruct-2506-mlx-fp16

VISIONConcurrent Unit Cost:2Model Size:24BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Jun 9, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The pa5haw/Mistral-Small-3.2-24B-Instruct-2506-mlx-fp16 model is a 24 billion parameter instruction-tuned language model, converted to the MLX format by pa5haw from the original Mistral-Small-3.2-24B-Instruct-2506 developed by Mistral AI. This model is optimized for efficient deployment and inference on Apple silicon using the MLX framework, making it suitable for local execution of instruction-following tasks. It maintains the original model's 32768 token context length, providing robust performance for complex prompts.

Loading preview...

Overview

This model, pa5haw/Mistral-Small-3.2-24B-Instruct-2506-mlx-fp16, is a 24 billion parameter instruction-tuned language model. It is a conversion of the original mistralai/Mistral-Small-3.2-24B-Instruct-2506 by Mistral AI into the MLX format, specifically designed for efficient execution on Apple silicon. The conversion was performed using mlx-lm version 0.31.2, ensuring compatibility and optimized performance within the MLX ecosystem.

Key Capabilities

  • Instruction Following: Inherits the instruction-tuned capabilities of the base Mistral-Small-3.2-24B-Instruct model, making it proficient at understanding and executing user commands.
  • MLX Optimization: Specifically formatted for the MLX framework, enabling high-performance inference on Apple devices.
  • Large Context Window: Supports a substantial context length of 32768 tokens, allowing for processing and generating longer, more complex texts.

Good For

  • Local Inference on Apple Silicon: Ideal for developers and users looking to run powerful instruction-tuned models directly on their Apple hardware (Macs with M-series chips) without relying on cloud services.
  • Prototyping and Development: Provides a robust local environment for developing applications that require a capable instruction-following LLM.
  • Applications Requiring Long Context: Suitable for tasks that benefit from a large context window, such as summarizing lengthy documents, complex code generation, or extended conversational AI.

Finetunes
1 models