Microsoft: Phi 4

Text input Text output
Author's Description

[Microsoft Research](/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...

Key Specifications
Cost
$$
Context
16K
Parameters
14B (Rumoured)
Released
Jan 09, 2025
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

Seed Frequency Penalty Top P Min P Response Format Temperature Stop Presence Penalty Max Tokens
Features

This model supports the following features:

Response Format
Performance Summary

Microsoft Phi-4, a 14-billion parameter model created on January 9, 2025, demonstrates a strong overall performance profile, particularly excelling in efficiency and reliability. It performs among the fastest models, ranking in the 63rd percentile for speed, and consistently offers highly competitive pricing, placing in the 85th percentile. Notably, Phi-4 exhibits exceptional reliability with a 100% success rate across all benchmarks, indicating minimal technical failures and consistent provision of usable responses. In terms of specific benchmark performance, Phi-4 achieved perfect accuracy in the Ethics (Baseline) test, making it the most accurate model at its price point and among models of comparable speed. It also showed strong performance in Hallucinations (Baseline) with 96.0% accuracy, effectively acknowledging uncertainty. While its General Knowledge (Baseline) accuracy was respectable at 96.8%, its performance in Email Classification (94.0% accuracy) and especially Reasoning (50.0% accuracy) suggests areas for improvement. Its strength lies in ethical reasoning and avoiding factual errors, while complex reasoning tasks present a notable weakness. Designed for complex reasoning and efficient operation in limited memory scenarios, Phi-4 is best suited for English language inputs.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $0.07
Completion $0.14

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
DeepInfra
DeepInfra | microsoft/phi-4 16K $0.07 / 1M tokens $0.14 / 1M tokens
Nebius
Nebius | microsoft/phi-4 16K $0.065 / 1M tokens $0.14 / 1M tokens
NextBit
NextBit | microsoft/phi-4 16K $0.065 / 1M tokens $0.14 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Strategy Free Executions Accuracy Cost Duration
Other Models by microsoft