Phi 4 Multimodal Instruct
Microsoft's Phi 4 Multimodal Instruct is a lightweight, open-source model that processes text, image, and audio inputs to generate text outputs, with a context window of 131,072 tokens. It is genuinely best at audio processing and speech-to-text tasks, and notably, it supports a wide range of languages across its modalities, including English, Chinese, and several European languages.
Input
Output
Context
131K
Max Output
4K
Parameters
5.6B
Technical Specifications
Model TypeAudio
Context Window131,072 tokens
Max Output Tokens4,096 tokens
Parameters5.6B
Release DateFeb 24, 2025
Training CutoffNot available
Licensemit
Open SourceYes
Input Modalities
Audio
Output Modalities
Text
Capabilities
Benchmarks
Artificial Analysis
4.4%
HLE
9.3%
AIME
31.5%
GPQA
11.0%
SciCode
69.3%
MATH 500
48.5%
MMLU Pro
13.1%
LiveCodeBench
4.5
Intelligence
14.2
Speed (tok/s)
0.421
TTFA (s)
0.421
TTFT (s)
Resources & Links
Estimated GPU Requirements for Phi 4 Multimodal Instruct
Estimates based on INT8 quantization. Actual requirements vary by framework and configuration.
Related Tools
Data sourced from official provider APIs and documentation
Last updated: Jun 23, 2026
Start building with the right model.
Automatically route workloads to the right model for every task, every time.