Jamba 1.5 Mini

Unknown Developer

This cutting-edge foundation model is part of the Jamba 1.5 family and excels at following instructions. It combines the strengths of SSM and Transformer architectures, resulting in exceptional long-context understanding, rapid processing speed, and high-quality output.

Model Specifications

Technical details and capabilities of Jamba 1.5 Mini

Core Specifications

52.0B Parameters

Model size and complexity

256.1K / 256.1K

Input / Output tokens

March 4, 2024

Knowledge cutoff date

August 21, 2024

Release date

Capabilities & License

Multimodal Support

Not Supported

Web Hydrated

No

License

Jamba Open Model License

Resources

Research Paper

https://arxiv.org/abs/2408.12570

API Reference

https://docs.ai21.com/reference/jamba-15-api-ref

Performance Insights

Check out how Jamba 1.5 Mini handles various AI tasks through comprehensive benchmark results.

90

68

45

23

0

85.7

ARC Challenge

85.7

(95%)

75.8

GSM-8K

75.8

(84%)

69.7

MMLU

69.7

(77%)

54.1

TruthfulQA

54.1

(60%)

46.1

Arena Hard

46.1

(51%)

42.5

MMLU-Pro

42.5

(47%)

42.4

Wild Bench

42.4

(47%)

32.3

GPQA

32.3

(36%)

ARC Challenge

GSM-8K

MMLU

TruthfulQA

Arena Hard

MMLU-Pro

Wild Bench

GPQA

Model Comparison

See how Jamba 1.5 Mini stacks up against other leading models across key performance metrics.

100

80

60

40

20

0

46.1

Arena Hard - Jamba 1.5 Mini

46.1

(46%)

37

Arena Hard - Phi-3.5-mini-instruct

37

(37%)

37.9

Arena Hard - Phi-3.5-MoE-instruct

37.9

(38%)

28.2

Arena Hard - Llama 3.1 8B Instruct

28.2

(28%)

65.4

Arena Hard - Jamba 1.5 Large

65.4

(65%)

69.7

MMLU - Jamba 1.5 Mini

69.7

(70%)

69

MMLU - Phi-3.5-mini-instruct

69

(69%)

78.9

MMLU - Phi-3.5-MoE-instruct

78.9

(79%)

69.4

MMLU - Llama 3.1 8B Instruct

69.4

(69%)

81.2

MMLU - Jamba 1.5 Large

81.2

(81%)

42.5

MMLU-Pro - Jamba 1.5 Mini

42.5

(43%)

47.4

MMLU-Pro - Phi-3.5-mini-instruct

47.4

(47%)

54.3

MMLU-Pro - Phi-3.5-MoE-instruct

54.3

(54%)

48.3

MMLU-Pro - Llama 3.1 8B Instruct

48.3

(48%)

53.5

MMLU-Pro - Jamba 1.5 Large

53.5

(54%)

32.3

GPQA - Jamba 1.5 Mini

32.3

(32%)

30.4

GPQA - Phi-3.5-mini-instruct

30.4

(30%)

36.8

GPQA - Phi-3.5-MoE-instruct

36.8

(37%)

30.4

GPQA - Llama 3.1 8B Instruct

30.4

(30%)

36.9

GPQA - Jamba 1.5 Large

36.9

(37%)

85.7

ARC Challenge - Jamba 1.5 Mini

85.7

(86%)

84.6

ARC Challenge - Phi-3.5-mini-instruct

84.6

(85%)

91

ARC Challenge - Phi-3.5-MoE-instruct

91

(91%)

83.4

ARC Challenge - Llama 3.1 8B Instruct

83.4

(83%)

93

ARC Challenge - Jamba 1.5 Large

93

(93%)

Arena Hard

MMLU

MMLU-Pro

GPQA

ARC Challenge

Jamba 1.5 Mini

Phi-3.5-mini-instruct

Phi-3.5-MoE-instruct

Llama 3.1 8B Instruct

Jamba 1.5 Large

Detailed Benchmarks

Dive deeper into Jamba 1.5 Mini's performance across specific task categories. Expand each section to see detailed metrics and comparisons.

Knowledge

MMLU

91.8%

71.8%

71.3%

Qwen2 7B Instruct

70.5%

69.8%

69.7%

Llama 3.1 8B Instruct

69.4%

69.2%

Phi-3.5-mini-instruct

69.0%

Llama 3.2 3B Instruct

63.4%

Current model

Other models

Avg (71.6%)

GPQA

87.7%

35.7%

33.3%

Llama 3.2 11B Instruct

32.8%

Llama 3.2 3B Instruct

32.8%

32.3%

30.8%

Llama 3.1 8B Instruct

30.4%

Phi-3.5-mini-instruct

30.4%

Qwen2 7B Instruct

25.3%

Current model

Other models

Avg (37.1%)

Hallucination

TruthfulQA

Phi-3.5-MoE-instruct

77.5%

Jamba 1.5 Large

58.3%

Qwen2.5 32B Instruct

57.8%

Qwen2 72B Instruct

54.8%

Qwen2.5-Coder 32B Instruct

54.2%

54.1%

Qwen2.5-Coder 7B Instruct

50.6%

Mistral NeMo Instruct

50.3%

Current model

Other models

Avg (57.2%)

Non categorized

Arena Hard

Mistral Small 3

87.6%

Ministral 8B Instruct

70.9%

Llama 3.3 70B Instruct

65.8%

Jamba 1.5 Large

65.4%

51.2%

46.1%

Phi-3.5-MoE-instruct

37.9%

Phi-3.5-mini-instruct

37.0%

33.1%

Llama 3.1 8B Instruct

28.2%

Current model

Other models

Avg (52.3%)

Wild Bench

Jamba 1.5 Large

48.5%

42.4%

Current model

Other models

Avg (45.5%)

MMLU-Pro

Jamba 1.5 Large

53.5%

51.0%

Qwen2.5-Coder 32B Instruct

50.4%

Llama 3.1 8B Instruct

48.3%

Phi-3.5-mini-instruct

47.4%

Qwen2 7B Instruct

44.1%

42.5%

Qwen2.5-Coder 7B Instruct

40.1%

Current model

Other models

Avg (47.2%)

ARC Challenge

Llama 3.1 405B Instruct

96.9%

Claude 3 Sonnet

93.2%

Jamba 1.5 Large

93.0%

Phi-3.5-MoE-instruct

91.0%

89.2%

85.7%

Phi-3.5-mini-instruct

84.6%

Llama 3.1 8B Instruct

83.4%

Ministral 8B Instruct

71.9%

Qwen2 72B Instruct

68.9%

Current model

Other models

Avg (85.8%)

GSM-8K

Jamba 1.5 Large

87.0%

75.8%

Current model

Other models

Avg (81.4%)

Providers Pricing Coming Soon

We're working on gathering comprehensive pricing data from all major providers for Jamba 1.5 Mini. 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