Model reference · open weights
Phi-4-reasoning is an open-weight language model from microsoft. Phi-4-reasoning (BF16) weighs 29.3 GB; the smallest configuration that runs it is L40S 48 GB.
For self-hosting, compare this build's requirements below with the GPU you plan to rent.
Summary of the microsoft/Phi-4-reasoning model card, 2026-10-01
What it is
| Released by | microsoft |
|---|---|
| Released | 2025-04-09 |
| Parameters | 14.7B |
| VRAM | 29.3 GB for the weights |
What it runs on
| Card | Requests at once | Context max | Memory | |
|---|---|---|---|---|
| 8K each | 32K each | |||
| RTX 3060 12 GB … RTX 4090 24 GB 4 smaller cards | ||||
| RTX 5090 32 GB | 4K | 31.0 GB | ||
| L40S 48 GB | 8 | 2 | all 32K | 44.0 GB |
| A100 80 GB | 28 | 7 | all 32K | 78.2 GB |
| H100 80 GB | 26 | 6 | all 32K | 78.1 GB |
| RTX PRO 6000 Blackwell 96 GB | 35 | 8 | all 32K | 93.8 GB |
| DGX Spark (GB10) 128 GB unified | 43 | 10 | all 32K | 107 GB |
| H200 141 GB | 61 | 15 | all 32K | 138 GB |
| B200 180 GB | 84 | 21 | all 32K | 176 GB |
| 2× RTX 4090 24 GB tensor parallel | 9 | 2 | all 32K | 23.4 GB a card |
| 2× RTX 3090 24 GB tensor parallel | 9 | 2 | all 32K | 23.4 GB a card |
| 2× RTX 5090 32 GB tensor parallel | 18 | 4 | all 32K | 31.0 GB a card |
| Requests at once | 8K tokens each | 32K tokens each |
|---|---|---|
| 1 | 31.8 GB | 36.8 GB |
| 5 | 38.5 GB | 63.6 GB |
| 8 | 43.5 GB | 83.8 GB |
| 16 | 56.9 GB | 137 GB |
| 32 | 83.8 GB | 245 GB |
| 64 | 137 GB | 460 GB |
One card, with vLLM's small-card settings.
From the model card
Phi-4-reasoning Technical Report
| Developers | Microsoft Research |
| Description | Phi-4-reasoning is a state-of-the-art open-weight reasoning model finetuned from Phi-4 using supervised fine-tuning on a dataset of chain-of-thought traces and reinforcement learning. The supervised fine-tuning dataset includes a blend of synthetic prompts and high-quality filtered data from public domain websites, focused on math, science, and coding skills as well as alignment data for safety and Responsible AI. The goal of this approach was to ensure that small capable models were trained with data focused on high quality and advanced reasoning. |
| Architecture | Base model same as previously released Phi-4, 14B parameters, dense decoder-only Transformer model |
| Inputs | Text, best suited for prompts in the chat format |
| Context length | 32k tokens |
| GPUs | 32 H100-80G |
| Training time | 2.5 days |
| Training data | 16B tokens, ~8.3B unique tokens |
| Outputs | Generated text in response to the input. Model responses have two sections, namely, a reasoning chain-of-thought block followed by a summarization block |
| Dates | January 2025 – April 2025 |
| Status | Static model trained on an offline dataset with cutoff dates of March 2025 and earlier for publicly available data |
| Release date | April 30, 2025 |
| License | MIT |
| Primary Use Cases | Our model is designed to accelerate research on language models, for use as a building block for generative AI powered features. It provides uses for general purpose AI systems and applications (primarily in English) which require:1. Memory/compute constrained environments.2. Latency bound scenarios.3. Reasoning and logic. |
| Out-of-Scope Use Cases | This model is designed and tested for math reasoning only. Our models are not specifically designed or evaluated for all downstream purposes. Developers should consider common limitations of language models as they select use cases, and evaluate and mitigate for accuracy, safety, and fairness before using within a specific downstream use case, particularly for high-risk scenarios. Developers should be aware of and adhere to applicable laws or regulations (including privacy, trade compliance laws, etc.) that are relevant to their use case, including the model’s focus on English. Review the Responsible AI Considerations section below for further guidance when choosing a use case. Nothing contained in this Model Card should be interpreted as or deemed a restriction or modification to the license the model is released under. |
[!IMPORTANT] To fully take advantage of the model's capabilities, inference must use
temperature=0.8,top_k=50,top_p=0.95, anddo_sample=True. For more complex queries, setmax_new_tokens=32768to allow for longer chain-of-thought (CoT).
Given the nature of the training data, always use ChatML template with the following system prompt for inference:
You are Phi, a language model trained by Microsoft to help users. Your role as an assistant involves thoroughly exploring questions through a systematic thinking process before providing the final precise and accurate solutions. This requires engaging in a comprehensive cycle of analysis, summarizing, exploration, reassessment, reflection, backtracing, and iteration to develop well-considered thinking process. Please structure your response into two main sections: Thought and Solution using the specified format: {Thought section} {Solution section}. In the Thought section, detail your reasoning process in steps. Each step should include detailed considerations such as analysing questions, summarizing relevant findings, brainstorming new ideas, verifying the accuracy of the current steps, refining any errors, and revisiting previous steps. In the Solution section, based on various attempts, explorations, and reflections from the Thought section, systematically present the final solution that you deem correct. The Solution section should be logical, accurate, and concise and detail necessary steps needed to reach the conclusion. Now, try to solve the following question through the above guidelines:
What is tQuoted from the model card on Hugging Face. The full card is behind the Hugging Face link above.