Gemma 3n E2B Instructed
Multimodal
Zero-eval
#2OpenAI MMLU
#3Global-MMLU
#3Codegolf v2.2
by Google
+
+
+
+
About
Gemma 3n E2B Instructed is a multimodal language model developed by Google. The model shows competitive results across 18 benchmarks. Notable strengths include HumanEval (66.5%), MMLU (60.1%), Global-MMLU-Lite (59.0%). 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
Performance Overview
Performance metrics and category breakdown
Overall Performance
18 benchmarks
Average Score
33.7%
Best Score
66.5%
High Performers (80%+)
0+
+
+
+
All Benchmark Results for Gemma 3n E2B Instructed
Complete list of benchmark scores with detailed information
| HumanEval | text | 0.67 | 66.5% | Self-reported | |
| MMLU | text | 0.60 | 60.1% | Self-reported | |
| Global-MMLU-Lite | text | 0.59 | 59.0% | Self-reported | |
| MBPP | text | 0.57 | 56.6% | Self-reported | |
| Global-MMLU | text | 0.55 | 55.1% | Self-reported | |
| MGSM | text | 0.53 | 53.1% | Self-reported | |
| WMT24++ | text | 0.43 | 42.7% | Self-reported | |
| MMLU-Pro | text | 0.41 | 40.5% | Self-reported | |
| Include | text | 0.39 | 38.6% | Self-reported | |
| HiddenMath | text | 0.28 | 27.7% | Self-reported |
Showing 1 to 10 of 18 benchmarks