Skip to main content
AITier

Qwen2.5 Instruct 72B

AlibabaQwenOpen WeightQwen · Commercial OK

Description

Qwen2.5-72B-Instruct is an instruction-tuned 72 billion parameter language model, part of the Qwen2.5 series. It is designed to follow instructions, generate long texts (over 8K tokens), understand structured data (e.g., tables), and generate structured outputs, especially JSON. The model supports multilingual capabilities across over 29 languages.

Release Date
2024-09-19
Parameters
72.7B
Context Length
33K
Modalities
text

Capability Radar

31
general
19
coding
29
reasoning
32
scienceest.
0
agents
0
multimodal

Science uses a reasoning proxy when dedicated science benchmarks are unavailable.

Rankings

Domain#RankScoreSource
Code Ranking334
21.0
AA
General Ranking285
38.0
AA
Math Reasoning270
29.0
AA
Science332
33.0
AA

Benchmark Scores (LLM Stats)

Biology

GPQA49.0%SR

Code

HumanEval86.6%SR
LiveCodeBench55.5%SR

Communication

MT-Bench0.94 / 100SR

Creativity

AlignBench81.6%SR
Arena Hard81.2%SR

Finance

MMLU-Pro71.1%SR

General

MBPP0.88 / 100SR
MMLU-Redux86.8%SR
IFEval84.1%SR
MultiPL-E75.1%SR
LiveBench52.3%SR

Math

GSM8k95.8%SR
MATH83.1%SR

AA Evaluation Indices

Intelligence Index
15.6
Math Index
14.0
Coding Index
11.9
Math 500
0.9
Mmlu Pro
0.7
Gpqa
0.5
Ifbench
0.4
Tau2
0.3
Livecodebench
0.3
Scicode
0.3
Lcr
0.2
Aime
0.2
Aime 25
0.1
Terminalbench Hard
0.0
Hle
0.0

LLM Stats Category Scores

Communication
90
Creativity
90
Roleplay
90
Structured Output
80
Writing
80
Instruction Following
80
Language
80
Math
80
Reasoning
80
Code
70
Finance
70
General
70
Healthcare
70
Legal
70
Biology
50
Chemistry
50
Physics
50

Pricing

Input Price$0.36 / 1M tokens
Output Price$0.4 / 1M tokens
Blended Price (3:1)$0.37 / 1M tokens

Speed

Tokens/sec54.7 tokens/s
Time to First Token1.08s
Time to Answer1.08s

Available Providers

(LS internal units)

No provider data available

External Sources

LLM Stats