跳转到主要内容
AITier

Phi-4

MicrosoftPhiOpen WeightMIT · Commercial OK

描述

phi-4 is a state-of-the-art open model built to excel at advanced reasoning, coding, and knowledge tasks. It leverages a blend of synthetic data, filtered web data, academic texts, and supervised fine-tuning for precision, alignment, and safety.

发布日期
2024-12-12
参数规模
14.7B
上下文长度
16K
支持模态
text

能力雷达图

28
general
17
coding
30
reasoning
36
science估算
0
agents
0
multimodal

Science 在缺少专门科学评测时使用推理能力代理估算。

排行榜排名

领域#排名分数来源
代码能力榜390
14.0
AA
通用能力榜415
23.0
AA
数学推理267
30.0
AA
推理能力25
83.0
LS
科学能力295
36.0
AA

基准测试分数 (LLM Stats)

Biology

GPQA56.1%自报

Code

HumanEval82.6%自报

Creativity

Arena Hard75.4%自报

Factuality

SimpleQA3.0%自报

Finance

MMLU84.8%自报
MMLU-Pro70.4%自报

General

IFEval63.0%自报
PhiBench56.2%自报
LiveBench47.6%自报

Math

MGSM80.6%自报
MATH80.4%自报
DROP75.5%自报

Reasoning

HumanEval+82.8%自报

AA 评测指数

Math Index
18.0
Coding Index
11.2
Intelligence Index
10.4
Math 500
0.8
Mmlu Pro
0.7
Gpqa
0.6
Scicode
0.3
Ifbench
0.2
Livecodebench
0.2
Aime 25
0.2
Aime
0.1
Hle
0.0
Terminalbench Hard
0.0
Lcr
0.0
Tau2
0.0

LLM Stats 分类评分

Writing
80
Code
80
Creativity
80
Finance
80
Healthcare
80
Language
80
Legal
80
Math
70
Reasoning
70
Structured Output
60
Biology
60
Chemistry
60
General
60
Instruction Following
60
Physics
60
Factuality
0

定价

输入价格$0.125 / 1M tokens
输出价格$0.5 / 1M tokens
混合价格(3:1)$0.219 / 1M tokens

速度

Tokens/秒38.5 tokens/s
首Token延迟0.51s
首回答延迟0.51s

可用提供商

(LS 内部计价单位)

暂无提供商数据

外部链接

LLM Stats