Phi-4 Reasoning

Small model with chain-of-thought reasoning

Phi-4 Reasoning is Microsoft's reasoning-focused variant, trained with reinforcement learning to produce step-by-step chain-of-thought traces. Despite its compact 14B size, it competes with much larger reasoning models on mathematics and science benchmarks.

Try Phi-4 Reasoning

Specifications

Context window33k tokens
Input price$0.150 / 1M tokens
Output price$0.600 / 1M tokens
Credits per query2 cr
Released2025-04
LicenseMIT

Features

Tool useJSON modeCodeExtended thinking

Benchmark Scores

What do these mean?

Sourced from official model cards and academic papers. Higher is better.

PhD-level science & reasoning

56%

Mathematical olympiad problems

78%

Chatbot Arena human preference ranking

Python coding — pass@1 rate

88%

Competition-level mathematics

91%

General knowledge across 57 academic domains

85.5%

Multi-turn instruction following (out of 10)

Real-world GitHub issue resolution rate