Gemma 3n E2B
Multimodal
Zero-eval
by Google
About
Gemma 3n E2B is a multimodal language model developed by Google. The model shows competitive results across 11 benchmarks. Notable strengths include PIQA (78.9%), BoolQ (76.4%), ARC-E (75.8%). As a multimodal model, it can process and understand text, images, and other input formats seamlessly. Released in 2025, it represents Google's latest advancement in AI technology.
Timeline
AnnouncedJun 26, 2025
ReleasedJun 26, 2025
Knowledge CutoffJun 1, 2024
Specifications
Training Tokens11.0T
Capabilities
Multimodal
License & Family
License
Proprietary
Benchmark Performance Overview
Performance metrics and category breakdown
Overall Performance
11 benchmarks
Average Score
58.6%
Best Score
78.9%
High Performers (80%+)
0Top Categories
reasoning
66.6%
general
54.1%
Benchmark Performance
Top benchmark scores with normalized values (0-100%)
Ranking Across Benchmarks
Position relative to other models on each benchmark
PIQA
Rank #8 of 9
#5Gemma 3n E2B Instructed LiteRT (Preview)
78.9%
#6Gemma 3n E4B
81.0%
#7Gemma 3n E4B Instructed LiteRT Preview
81.0%
#8Gemma 3n E2B
78.9%
#9Phi 4 Mini
77.6%
BoolQ
Rank #8 of 9
#5Phi-3.5-mini-instruct
78.0%
#6Phi 4 Mini
81.2%
#7Gemma 3n E4B Instructed LiteRT Preview
81.6%
#8Gemma 3n E2B
76.4%
#9Gemma 3n E2B Instructed LiteRT (Preview)
76.4%
ARC-E
Rank #6 of 6
#3Gemma 3n E2B Instructed LiteRT (Preview)
75.8%
#4Gemma 3n E4B Instructed LiteRT Preview
81.6%
#5Gemma 3n E4B
81.6%
#6Gemma 3n E2B
75.8%
HellaSwag
Rank #21 of 24
#18Gemma 3n E2B Instructed LiteRT (Preview)
72.2%
#19Qwen2.5-Coder 7B Instruct
76.8%
#20Gemma 3n E4B
78.6%
#21Gemma 3n E2B
72.2%
#22Llama 3.2 3B Instruct
69.8%
#23Phi-3.5-mini-instruct
69.4%
#24Phi 4 Mini
69.1%
Winogrande
Rank #18 of 19
#15Phi 4 Mini
67.0%
#16Phi-3.5-mini-instruct
68.5%
#17Gemma 3n E4B Instructed LiteRT Preview
71.7%
#18Gemma 3n E2B
66.8%
#19Gemma 3n E2B Instructed LiteRT (Preview)
66.8%
All Benchmark Results for Gemma 3n E2B
Complete list of benchmark scores with detailed information
PIQA PIQA benchmark | general | text | 0.79 | 78.9% | Self-reported |
BoolQ BoolQ benchmark | general | text | 0.76 | 76.4% | Self-reported |
ARC-E ARC-E benchmark | reasoning | text | 0.76 | 75.8% | Self-reported |
HellaSwag HellaSwag benchmark | reasoning | text | 0.72 | 72.2% | Self-reported |
Winogrande Winogrande benchmark | reasoning | text | 0.67 | 66.8% | Self-reported |
TriviaQA TriviaQA benchmark | general | text | 0.61 | 60.8% | Self-reported |
DROP DROP benchmark | general | text | 0.54 | 53.9% | Self-reported |
ARC-C ARC-C benchmark | reasoning | text | 0.52 | 51.7% | Self-reported |
Social IQa Social IQa benchmark | general | text | 0.49 | 48.8% | Self-reported |
BIG-Bench Hard BIG-Bench Hard benchmark | general | text | 0.44 | 44.3% | Self-reported |
Showing 1 to 10 of 11 benchmarks