Phi 4

Microsoft

Phi-4 is designed to perform well in complex reasoning tasks and can operate efficiently in situations with limited memory or where quick responses are needed. At 14 billion parameters, it was trained on a mix of high-quality synthetic datasets, data from curated websites, and academic materials. It has undergone careful improvement to follow instructions accurately and maintain strong safety standards. It works best with English language inputs.

Try Now

Technical Specifications

Context Window

16,384 tokens

Max Output

16,384 tokens

Pricing

Token Costs (per 1M tokens)

Cache Miss Input

$0.06

Non-Reasoning Output

$0.14

Legacy

Made legacy on

Reason

14B model; superseded by larger frontier models

Recommended Replacement

Qwen3.6 Plus