Microsoft Phi-4-reasoning-plus

Phi-4-reasoning-plus is a state-of-the-art open-weight reasoning large language model developed by Microsoft. Finetuned from the base Phi-4 model with a mix of supervised fine-tuning on chain-of-thought traces and reinforcement learning, it is optimized for advanced reasoning tasks in math, science, and coding. The model features a dense 14B parameter decoder-only Transformer architecture with a 32k token context length and produces responses with a reasoning chain-of-thought followed by a summarization. It is intended for research and generative AI applications in constrained memory/latency settings.