Qwen2.5 7B Instruct

Qwen

Qwen2.5-7B-Instruct is a 7 billion parameter language model finely tuned to master instruction following. It demonstrates exceptional proficiency in creating extended content (over 8,000 tokens), interpreting structured information, and producing structured output formats such as JSON. This model offers improved performance in mathematical reasoning, coding tasks, and multilingual applications, supporting over 29 languages, including Chinese, English, French, and Spanish.

Model Specifications

Technical details and capabilities of Qwen2.5 7B Instruct

Core Specifications

7.6B Parameters

Model size and complexity

18000.0B Training Tokens

Amount of data used in training

131.1K / 8.2K

Input / Output tokens

September 18, 2024

Release date

Capabilities & License

Multimodal Support

Not Supported

Web Hydrated

No

License

apache-2.0

Resources

Research Paper

https://arxiv.org/abs/2407.10671

API Reference

https://www.alibabacloud.com/help/en/model-studio/developer-reference/use-qwen-by-calling-api

Code Repository

https://github.com/QwenLM/Qwen2.5

Performance Insights

Check out how Qwen2.5 7B Instruct handles various AI tasks through comprehensive benchmark results.

100

75

50

25

0

91.6

GSM8K

91.6

(92%)

87.5

MT-bench

87.5

(88%)

84.8

HumanEval

84.8

(85%)

79.2

MBPP

79.2

(79%)

75.5

MATH

75.5

(76%)

75.4

MMLU-Redux

75.4

(75%)

73.3

AlignBench

73.3

(73%)

71.2

IFEval

71.2

(71%)

70.4

MultiPL-E

70.4

(70%)

56.3

MMLU-Pro

56.3

(56%)

52

Arena-Hard

52

(52%)

36.4

GPQA

36.4

(36%)

35.9

LiveBench

35.9

(36%)

28.7

LiveCodeBench

28.7

(29%)

GSM8K

MT-bench

HumanEval

MBPP

MATH

MMLU-Redux

AlignBench

IFEval

MultiPL-E

MMLU-Pro

Arena-Hard

GPQA

LiveBench

LiveCodeBench

Model Comparison

See how Qwen2.5 7B Instruct stacks up against other leading models across key performance metrics.

100

80

60

40

20

0

56.3

MMLU-Pro - Qwen2.5 7B Instruct

56.3

(56%)

63.7

MMLU-Pro - Qwen2.5 14B Instruct

63.7

(64%)

64.4

MMLU-Pro - Qwen2 72B Instruct

64.4

(64%)

69

MMLU-Pro - Qwen2.5 32B Instruct

69

(69%)

71.1

MMLU-Pro - Qwen2.5 72B Instruct

71.1

(71%)

54.3

MMLU-Pro - Phi-3.5-MoE-instruct

54.3

(54%)

75.5

MATH - Qwen2.5 7B Instruct

75.5

(76%)

80

MATH - Qwen2.5 14B Instruct

80

(80%)

59.7

MATH - Qwen2 72B Instruct

59.7

(60%)

83.1

MATH - Qwen2.5 32B Instruct

83.1

(83%)

83.1

MATH - Qwen2.5 72B Instruct

83.1

(83%)

59.5

MATH - Phi-3.5-MoE-instruct

59.5

(60%)

91.6

GSM8K - Qwen2.5 7B Instruct

91.6

(92%)

94.8

GSM8K - Qwen2.5 14B Instruct

94.8

(95%)

91.1

GSM8K - Qwen2 72B Instruct

91.1

(91%)

95.9

GSM8K - Qwen2.5 32B Instruct

95.9

(96%)

95.8

GSM8K - Qwen2.5 72B Instruct

95.8

(96%)

88.7

GSM8K - Phi-3.5-MoE-instruct

88.7

(89%)

84.8

HumanEval - Qwen2.5 7B Instruct

84.8

(85%)

83.5

HumanEval - Qwen2.5 14B Instruct

83.5

(84%)

86

HumanEval - Qwen2 72B Instruct

86

(86%)

88.4

HumanEval - Qwen2.5 32B Instruct

88.4

(88%)

86.6

HumanEval - Qwen2.5 72B Instruct

86.6

(87%)

70.7

HumanEval - Phi-3.5-MoE-instruct

70.7

(71%)

36.4

GPQA - Qwen2.5 7B Instruct

36.4

(36%)

45.5

GPQA - Qwen2.5 14B Instruct

45.5

(46%)

42.4

GPQA - Qwen2 72B Instruct

42.4

(42%)

49.5

GPQA - Qwen2.5 32B Instruct

49.5

(50%)

49

GPQA - Qwen2.5 72B Instruct

49

(49%)

36.8

GPQA - Phi-3.5-MoE-instruct

36.8

(37%)

MMLU-Pro

MATH

GSM8K

HumanEval

GPQA

Qwen2.5 7B Instruct

Qwen2.5 14B Instruct

Qwen2 72B Instruct

Qwen2.5 32B Instruct

Qwen2.5 72B Instruct

Phi-3.5-MoE-instruct

Detailed Benchmarks

Dive deeper into Qwen2.5 7B Instruct's performance across specific task categories. Expand each section to see detailed metrics and comparisons.

Math

GSM8K

97.1%

Qwen2.5 14B Instruct

94.8%

Mistral Large 2

93.0%

92.3%

Claude 3 Sonnet

92.3%

Qwen2.5 7B Instruct

91.6%

Llama 3.1 Nemotron 70B Instruct

91.4%

Qwen2.5-Coder 32B Instruct

91.1%

Qwen2 72B Instruct

91.1%

68.6%

Current model

Other models

Avg (90.3%)

Coding

HumanEval

Claude 3.5 Sonnet

93.7%

85.7%

85.4%

84.9%

Mistral Small 3

84.8%

Qwen2.5 7B Instruct

84.8%

84.1%

Qwen2.5 14B Instruct

83.5%

82.6%

Ministral 8B Instruct

34.8%

Current model

Other models

Avg (80.4%)

MBPP

Qwen2.5-Coder 32B Instruct

90.2%

Qwen2.5-Coder 7B Instruct

83.5%

Qwen2.5 14B Instruct

82.0%

Phi-3.5-MoE-instruct

80.8%

Qwen2 72B Instruct

80.2%

Qwen2.5 7B Instruct

79.2%

78.2%

74.4%

Phi-3.5-mini-instruct

69.6%

52.4%

Current model

Other models

Avg (77.0%)

LiveCodeBench

80.0%

39.0%

37.6%

Gemini 2.0 Flash

35.1%

Qwen2.5-Coder 32B Instruct

31.4%

Qwen2.5 7B Instruct

28.7%

Qwen2 7B Instruct

26.6%

Qwen2.5-Coder 7B Instruct

18.2%

Current model

Other models

Avg (37.1%)

Knowledge

MATH

97.9%

Llama 3.3 70B Instruct

77.0%

76.6%

76.6%

76.1%

Qwen2.5 7B Instruct

75.5%

74.7%

Llama 3.1 405B Instruct

73.8%

73.3%

32.6%

Current model

Other models

Avg (73.4%)

GPQA

87.7%

40.0%

Gemini 1.5 Flash 8B

38.4%

Jamba 1.5 Large

36.9%

Phi-3.5-MoE-instruct

36.8%

Qwen2.5 7B Instruct

36.4%

35.9%

35.7%

33.3%

Qwen2 7B Instruct

25.3%

Current model

Other models

Avg (40.6%)

Non categorized

MMLU-Pro

84.0%

Qwen2 72B Instruct

64.4%

Qwen2.5 14B Instruct

63.7%

Gemini 1.5 Flash 8B

58.7%

Claude 3 Sonnet

56.8%

Qwen2.5 7B Instruct

56.3%

Phi-3.5-MoE-instruct

54.3%

Jamba 1.5 Large

53.5%

51.0%

Qwen2.5-Coder 7B Instruct

40.1%

Current model

Other models

Avg (58.3%)

MMLU-Redux

92.9%

89.1%

Qwen2.5 72B Instruct

86.8%

Qwen2.5 32B Instruct

83.9%

Qwen2.5 14B Instruct

80.0%

Qwen2.5-Coder 32B Instruct

77.5%

Qwen2.5 7B Instruct

75.4%

Qwen2.5-Coder 7B Instruct

66.6%

Current model

Other models

Avg (81.5%)

MultiPL-E

Qwen2.5 32B Instruct

75.4%

Qwen2.5 72B Instruct

75.1%

Qwen2.5 14B Instruct

72.8%

Qwen2.5 7B Instruct

70.4%

Qwen2 72B Instruct

69.2%

Qwen2 7B Instruct

59.1%

Current model

Other models

Avg (70.3%)

IFEval

Claude 3.7 Sonnet

93.2%

83.3%

Mistral Small 3

82.9%

Llama 3.1 8B Instruct

80.4%

Llama 3.2 3B Instruct

77.4%

Qwen2.5 7B Instruct

71.2%

63.0%

61.3%

Current model

Other models

Avg (76.6%)

Arena-Hard

Qwen2.5 72B Instruct

81.2%

Qwen2.5 7B Instruct

52.0%

Current model

Other models

Avg (66.6%)

AlignBench

Qwen2.5 72B Instruct

81.6%

80.4%

Qwen2.5 7B Instruct

73.3%

Qwen2 7B Instruct

72.1%

Current model

Other models

Avg (76.9%)

MT-bench

Qwen2.5 72B Instruct

93.5%

Qwen2.5 7B Instruct

87.5%

Current model

Other models

Avg (90.5%)

LiveBench

73.1%

67.0%

52.3%

Qwen2.5 72B Instruct

52.3%

47.6%

Qwen2.5 7B Instruct

35.9%

Current model

Other models

Avg (54.7%)

Providers Pricing Coming Soon

We're working on gathering comprehensive pricing data from all major providers for Qwen2.5 7B Instruct. Compare costs across platforms to find the best pricing for your use case.

OpenAI

Anthropic

Google

Mistral AI

Cohere

Share your feedback

Hi, I'm Charlie Palars, the founder of Deepranking.ai. I'm always looking for ways to improve the site and make it more useful for you. You can write me through this form or directly through X at @palarsio.

Your feedback helps us improve our service

Stay Ahead with AI Updates

Get insights on Gemini Pro 2.5, Sonnet 3.7 and more top AI models