%20--%3e%3cdefs%3e%3cstyle%3e%20.st0%20{%20fill:%20%23061b40;%20}%20.st1%20{%20fill:%20%23306af1;%20}%20.st2%20{%20fill:%20%235ce5cf;%20}%20%3c/style%3e%3c/defs%3e%3cg%3e%3cpath%20class='st0'%20d='M55,10.5h9v3h-9v9h12V7.5h-12v3ZM64,19.5h-6v-3h6v3Z'/%3e%3cpolygon%20class='st0'%20points='69%2016.5%2078%2016.5%2078%2019.5%2069%2019.5%2069%2022.5%2081%2022.5%2081%2013.5%2072%2013.5%2072%2010.5%2081%2010.5%2081%207.5%2069%207.5%2069%2016.5'/%3e%3cpolygon%20class='st0'%20points='95%2010.5%2095%207.5%2083%207.5%2083%2022.5%2095%2022.5%2095%2019.5%2086%2019.5%2086%2016.5%2095%2016.5%2095%2013.5%2086%2013.5%2086%2010.5%2095%2010.5'/%3e%3cpath%20class='st0'%20d='M40,1.5v21h11.6l1.4-1.4v-7.6h0c0,0-1.4-1.5-1.4-1.5l1.4-1.4V2.9l-1.4-1.4h-11.6ZM50,19.5h-7v-6h7v6ZM50,10.5h-7v-6h7v6Z'/%3e%3c/g%3e%3cpath%20class='st1'%20d='M23.1,24L14.7,4.8l-4.9,11.2h3.8l-1.8,4H3.3L12.1,0H2C.9,0,0,.9,0,2v20c0,1.1.9,2,2,2h21.1Z'/%3e%3cpath%20class='st2'%20d='M34,0h-16.8l10.6,24h6.2c1.1,0,2-.9,2-2V2C36,.9,35.1,0,34,0ZM32.5,20h-4V4h4v16Z'/%3e%3c/svg%3e)
Phi 4
Phi-4 is a state-of-the-art open-source model designed to excel in advanced reasoning, coding, and knowledge tasks. It combines synthetic data, curated web data, and academic texts, and undergoes supervised fine-tuning to enhance precision, alignment, and security.
Intelligence(Relatively Weak)
Speed(Slow)
Input Supported Modalities
No
Is Reasoning Model
16,000
Context Window
16,000
Maximum Output Tokens
2024-06-01
Knowledge Cutoff
Pricing
￥0.5 /M tokens
Input
￥1.01 /M tokens
Output
￥1.57 /M tokens
Blended Price
Quick Simple Comparison
Phi-4 Mini Instruct
Phi-4 Multimodal Instruct
￥0.05
Phi-3 Medium Instruct 14B
￥0.1
Basic Parameters
Phi-4Technical Parameters
Parameter Count
14,700.0M
Context Length
16.00k tokens
Training Data Cutoff
2024-06-01
Open Source Category
Open Weights (Permissive License)
Multimodal Support
Text Only
Throughput
33
Release Date
2024-12-12
Response Speed
25.212,677 tokens/s
Benchmark Scores
Below is the performance of Phi-4 in various standard benchmark tests. These tests evaluate the model's capabilities in different tasks and domains.
Intelligence Index
40.22
Large Language Model Intelligence Level
Coding Index
24.56
Indicator of AI model performance on coding tasks
Math Index
47.67
Capability indicator in solving mathematical problems, mathematical reasoning, or performing math-related tasks
MMLU Pro
71.4
Massive Multitask Multimodal Understanding - Testing understanding of text, images, audio, and video
GPQA
57.5
Graduate Physics Questions Assessment - Testing advanced physics knowledge with diamond science-level questions
HLE
4.1
The model's comprehensive average score on the Hugging Face Open LLM Leaderboard
LiveCodeBench
23.1
Specific evaluation focused on assessing large language models' ability in real-world code writing and solving programming competition problems
SciCode
26
The model's capability in code generation for scientific computing or specific scientific domains
HumanEval
86.7
Score achieved by the AI model on the specific HumanEval benchmark test set
Math 500 Score
81
Score on the first 500 larger, more well-known mathematical benchmark tests
AIME Score
14.3
An indicator measuring an AI model's ability to solve high-difficulty mathematical competition problems (specifically AIME level)
GPT 5 Mini
openai
￥1.8
Input tokens/million
￥14.4
Output tokens/million
400k
Context Length
GPT 5 Standard
openai
￥63
Input tokens/million
￥504
Output tokens/million
400k
Context Length
GPT 5 Nano
openai
￥0.36
Input tokens/million
￥2.88
Output tokens/million
400k
Context Length
GPT 5
openai
￥9
Input tokens/million
￥72
Output tokens/million
400k
Context Length
GLM 4.5
chatglm
￥0.43
Input tokens/million
￥1.01
Output tokens/million
131k
Context Length
Gemini 2.0 Flash Lite (Preview)
google
￥0.58
Input tokens/million
￥2.16
Output tokens/million
1M
Context Length
Gemini 1.0 Pro
google
￥3.6
Input tokens/million
￥10.8
Output tokens/million
33k
Context Length
Qwen2.5 Coder Instruct 32B
alibaba
￥0.65
Input tokens/million
￥0.65
Output tokens/million
131k
Context Length