Qwen 2.5 Coder 32B Instruct is an LLM designed for code generation, reasoning, and fixing. It excels in coding tasks, making it a powerful tool for developers. The model is based on a transformer architecture with advanced techniques like RoPE and SwiGLU.

Metric	Value
Parameter Count	32.5 billion
Mixture of Experts	No
Context Length	128,000 tokens
Multilingual	Unknown
Quantized*	Unknown

*Quantization is specific to the inference provider and the model may be offered with different quantization levels by other providers.