FlatFootInternational/Qwen2.5-Coder-14B-Instruct-MLX

TEXT GENERATIONConcurrency Cost:1Model Size:14.8BQuant:FP8Ctx Length:32kPublished:Feb 6, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

FlatFootInternational/Qwen2.5-Coder-14B-Instruct-MLX is a 14.8 billion parameter instruction-tuned language model, converted to MLX format from Qwen/Qwen2.5-Coder-14B-Instruct. This model is specifically designed for code generation and understanding tasks, leveraging its large parameter count and instruction-tuning for enhanced performance in programming-related applications. It offers a substantial context length of 131072 tokens, making it suitable for handling extensive codebases and complex coding prompts. Its primary use case is facilitating advanced code generation, debugging, and development assistance within the MLX ecosystem.

Loading preview...

Overview

FlatFootInternational/Qwen2.5-Coder-14B-Instruct-MLX is a 14.8 billion parameter instruction-tuned language model, converted for use with the MLX framework. This model originates from the Qwen/Qwen2.5-Coder-14B-Instruct base, indicating its specialization in coding tasks. The conversion to MLX format was performed using mlx-lm version 0.30.6, enabling efficient deployment and inference on Apple silicon.

Key Capabilities

  • Code Generation: Optimized for generating high-quality code across various programming languages.
  • Instruction Following: Designed to accurately follow complex coding instructions and prompts.
  • Large Context Window: Features a substantial context length of 131072 tokens, allowing it to process and generate code for large projects or intricate problem descriptions.
  • MLX Compatibility: Fully compatible with the MLX ecosystem, providing seamless integration for developers working with Apple hardware.

Good For

  • Software Development: Assisting developers with writing, debugging, and refactoring code.
  • Code Understanding: Analyzing and explaining existing code snippets or functions.
  • Prototyping: Rapidly generating code prototypes based on natural language descriptions.
  • Educational Tools: Serving as a powerful backend for coding tutorials or learning platforms that require code generation capabilities.