lmstudio-community/magistral-small-2506-mlx-bf16
TEXT GENERATIONConcurrency Cost:2Model Size:24BQuant:FP8Ctx Length:32kPublished:Jun 10, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Warm
The magistral-small model by mistral-ai is a 24 billion parameter language model, specifically the bfloat16 version, optimized for Apple Silicon. This model is provided as an MLX quantization, making it suitable for efficient local inference on Apple hardware. It leverages the MLX framework for performance, targeting developers working within the Apple ecosystem.
Loading preview...
Model Overview
The magistral-small model, developed by mistral-ai, is a 24 billion parameter language model. This particular version is an MLX quantization of the original bfloat16 magistral-small model, specifically optimized for efficient execution on Apple Silicon.
Key Characteristics
- Creator: mistral-ai
- Original Model:
magistral-small(bfloat16 version) - Quantization: MLX, provided by the LM Studio team using
mlx_lm. - Hardware Optimization: Designed for Apple Silicon, leveraging the MLX framework developed by the Apple Machine Learning Research team.
Use Cases
This model is particularly well-suited for developers and researchers who:
- Require a powerful 24 billion parameter language model.
- Are working within the Apple ecosystem and need models optimized for Apple Silicon.
- Seek efficient local inference capabilities on their Apple hardware, benefiting from the MLX framework's performance advantages.