Skip to content

Phi-4-multimodal-instruct

A multimodal model for text, image and audio input

Visit official site

What Phi-4-multimodal-instruct is

The model is described as a Microsoft multimodal foundation model that accepts text, images and audio and returns text.

What you can do with it

  • Run automatic speech recognition and translation
  • Answer questions about multiple images
  • Work with speech summaries and speech QA

Application scenarios

  • A voice assistant gives multilingual speech answers
  • Education uses visual and audio input for math and science
  • Content creation writes text from images or audio

Pros & cons

✓ What we like

  • Multiple image input supports chart understanding
  • Speech summary and speech QA are listed

! What to watch out for

  • Output is text only
  • Images are converted to a supported format first

FAQ

What context length is named?

The material states a 128K context length.

Which languages are listed?

The material names English, Chinese and French for text, and English, Chinese and German for audio.

Last reviewed: 2026-11-15

More LLM API platform tools

View all →

How we review