Phi-4-mini-reasoning

Phi-4-mini-reasoning is a lightweight open model from the Phi-4 family, built on synthetic, reasoning-dense data and further finetuned for advanced math reasoning. It supports a 128K token context length and is ideal for multi-step, logic-intensive mathematical problem-solving in memory- or compute-constrained environments. Use cases include formal proof generation, symbolic computation, and advanced word problems, with a focus on deep analytical thinking.

Note: This model is designed and tested for math reasoning only. Evaluate and mitigate for accuracy, safety, and fairness before deploying in downstream or high-risk scenarios. Refer to the model license for usage terms.