Phi-4 Mini Reasoning 3.8B

Overview

Microsoft's 3.8B Phi-4 Mini variant trained on R1-style reasoning traces under MIT. AIME24 57.5 and MATH-500 94.6 — remarkable math chops for the size.

When to pick this model

• Math and STEM reasoning on a laptop
• Educational tutoring apps under MIT
• Research into small-model reasoning distillation
• Battery-constrained reasoning workloads

VRAM requirements by quantization

Quantization	VRAM required
Q4_K_M (recommended)	10 GB
Q5_K_M	12 GB
Q8_0	18 GB
FP16 (no quantization)	33 GB

VRAM figures include model weights plus a typical 8k KV cache and ~600 MB runtime overhead (Ollama / llama.cpp baseline). Add headroom for higher context lengths.

Published benchmark scores

Benchmark	Score
AIME 2024	57.5
MATH-500	94.6

Scores published by the model author or aggregated from public leaderboards. Re-measured monthly by our editorial team.

Strengths

• AIME24 57.5 — exceptional for 3.8B
• MATH-500 94.6 nearly matches frontier models
• Fits comfortably on any laptop
• MIT license

Limitations

• English-first
• Verbose CoT typical of reasoning models
• Outside math, quality trails the base Phi-4 Mini

Architecture & training

Architecture: Dense 3.8B · trained on R1 traces

Training: Synthetic reasoning distillation.

Quick start

ollama run phi4-mini-reasoning:3.8b

Or use the open-source MCP server to query this model from Claude Desktop, Cursor, or any MCP-compatible client.