Microsoft continues so as to add to the dialog by unveiling its latest fashions, Phi-4-reasoning, Phi-4-reasoning-plus, and Phi-4-mini-reasoning.
A brand new period of AI
One 12 months in the past, Microsoft launched small language fashions (SLMs) to prospects with the discharge of Phi-3 on Azure AI Foundry, leveraging analysis on SLMs to increase the vary of environment friendly AI fashions and instruments obtainable to prospects.
Today, we’re excited to introduce Phi-4-reasoning, Phi-4-reasoning-plus, and Phi-4-mini-reasoning—marking a brand new period for small language fashions and as soon as once more redefining what is feasible with small and environment friendly AI.
Reasoning fashions, the following step ahead
Reasoning fashions are educated to leverage inference-time scaling to carry out complicated duties that demand multi-step decomposition and inner reflection. They excel in mathematical reasoning and are rising because the spine of agentic purposes with complicated, multi-faceted duties. Such capabilities are usually discovered solely in massive frontier fashions. Phi-reasoning fashions introduce a brand new class of small language fashions. Using distillation, reinforcement studying, and high-quality information, these fashions steadiness dimension and efficiency. They are sufficiently small for low-latency environments but keep robust reasoning capabilities that rival a lot larger fashions. This mix permits even resource-limited gadgets to carry out complicated reasoning duties effectively.
Phi-4-reasoning and Phi-4-reasoning-plus
Phi-4-reasoning is a 14-billion parameter open-weight reasoning mannequin that rivals a lot bigger fashions on complicated reasoning duties. Trained by way of supervised fine-tuning of Phi-4 on rigorously curated reasoning demonstrations from OpenAI o3-mini, Phi-4-reasoning generates detailed reasoning chains that successfully leverage extra inference-time compute. The mannequin demonstrates that meticulous information curation and high-quality artificial datasets enable smaller fashions to compete with bigger counterparts.
Phi-4-reasoning-plus builds upon Phi-4-reasoning capabilities, additional educated with reinforcement studying to make the most of extra inference-time compute, utilizing 1.5x extra tokens than Phi-4-reasoning, to ship greater accuracy.
Despite their considerably smaller dimension, each fashions obtain higher efficiency than OpenAI o1-mini and DeepSeek-R1-Distill-Llama-70B at most benchmarks, together with mathematical reasoning and Ph.D. stage science questions. They obtain efficiency higher than the complete DeepSeek-R1 mannequin (with 671-billion parameters) on the AIME 2025 take a look at, the 2025 qualifier for the USA Math Olympiad. Both fashions can be found on Azure AI Foundry and HuggingFace, right here and right here.
Phi-4-reasoning fashions introduce a significant enchancment over Phi-4, surpass bigger fashions like DeepSeek-R1-Distill-70B and strategy Deep-Seek-R1 throughout varied reasoning and common capabilities, together with math, coding, algorithmic drawback fixing, and planning. The technical report supplies in depth quantitative proof of those enhancements via numerous reasoning duties.
Phi-4-mini-reasoning
Phi-4-mini-reasoning is designed to fulfill the demand for a compact reasoning mannequin. This transformer-based language mannequin is optimized for mathematical reasoning, offering high-quality, step-by-step drawback fixing in environments with constrained computing or latency. Fine-tuned with artificial information generated by Deepseek-R1 mannequin, Phi-4-mini-reasoning balances effectivity with superior reasoning capability. It’s supreme for academic purposes, embedded tutoring, and light-weight deployment on edge or cellular methods, and is educated on over a million numerous math issues spanning a number of ranges of issue from center faculty to Ph.D. stage. Try out the mannequin on Azure AI Foundry or HuggingFace at present.
For extra details about the mannequin, learn the technical report that gives extra quantitative insights.
Phi’s evolution over the past 12 months has regularly pushed this envelope of high quality vs. dimension, increasing the household with new options to deal with numerous wants. Across the size of Windows 11 gadgets, these fashions can be found to run regionally on CPUs and GPUs.
As Windows works in the direction of creating a brand new sort of PC, Phi fashions have grow to be an integral a part of Copilot+ PCs with the NPU-optimized Phi Silica variant. This extremely environment friendly and OS-managed model of Phi is designed to be preloaded in reminiscence, and obtainable with blazing quick time to first token responses, and energy environment friendly token throughput so it may be concurrently invoked with different purposes working in your PC.
It is utilized in core experiences like Click to Do, offering helpful textual content intelligence instruments for any content material in your display screen, and is offered as developer APIs to be readily built-in into purposes—already being utilized in a number of productiveness purposes like Outlook, providing its Copilot abstract options offline. These small however mighty fashions have already been optimized and built-in for use throughout a number of purposes throughout the breadth of our PC ecosystem. The Phi-4-reasoning and Phi-4-mini-reasoning fashions leverage the low-bit optimizations for Phi Silica and will likely be obtainable to run quickly on Copilot+ PC NPUs.
Safety and Microsoft’s strategy to accountable AI
At Microsoft, accountable AI is a basic precept guiding the event and deployment of AI methods, together with our Phi fashions. Phi fashions are developed in accordance with Microsoft AI ideas: accountability, transparency, equity, reliability and security, privateness and safety, and inclusiveness.
The Phi household of fashions has adopted a sturdy security post-training strategy, leveraging a mixture of Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), and Reinforcement Learning from Human Feedback (RLHF) strategies. These strategies make the most of varied datasets, together with publicly obtainable datasets targeted on helpfulness and harmlessness, in addition to varied safety-related questions and solutions. While the Phi household of fashions is designed to carry out a variety of duties successfully, it is very important acknowledge that every one AI fashions could exhibit limitations. To higher perceive these limitations and the measures in place to deal with them, please consult with the mannequin playing cards beneath, which offer detailed info on accountable AI practices and tips.