Benchmark Almanac

Gemma 3n E2B

Gemma 3n E2B.

Google DeepMind Released June 26th, 2025 Family Gemma 3n The lab's model card Compare with others
14
Benchmarks with a score
0
Ranked first, out of five or more models
0
In the top three
Google DeepMind Jun 2025

Gemma 3n E2B

Open weights Small

No Epoch Capabilities Index score yet
  • MMAU-Pro 35.4% #11 /12
  • Angry Tweets 55.2% #59 /149
  • FoQA 69% #29 /91

How does it do against its rivals?

Gemma 3n E2B beats gpt-oss-20b on 7 of the 10 benchmarks they share. Qwen3-8B does best against it, ahead on 6 of 9.

  1. gpt-oss-20b 7 ahead, 3 behind Compare
  2. Qwen3-1.7B 6 ahead, 3 behind, 1 tied Compare
  3. Qwen3-4B-Thinking-2507 5 ahead, 3 behind Compare
  4. Qwen3-4B 5 ahead, 5 behind Compare
  5. OLMo 2 32B Instruct 4 ahead, 6 behind Compare
  6. Qwen3-8B 3 ahead, 6 behind Compare

Sources: Official leaderboards, Epoch AI (CC BY 4.0) and labs’ model cards

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Knowledge and language

BenchmarkScoreRankSource
FoQA 69% #29 of 91 Official leaderboard
Biblically Informed Bot Learning Evaluation 42.8% #6 of 18 Official leaderboard
Angry Tweets 55.2% #59 of 149 Official leaderboard
FoSent 50.2% #64 of 147 Official leaderboard
Dutch Book Reviews Dataset 89% #49 of 105 Official leaderboard
GermanQuAD 54% #47 of 88 Official leaderboard
GuiaCat 47.6% #22 of 35 Official leaderboard
BLUEX 47.3% #120 of 177 Official leaderboard
Open Portuguese LLM Leaderboard 56.3 #126 of 175 Official leaderboard
SEND Pedagogy Benchmark 60.5% #207 of 230 Official leaderboard
Pedagogy Benchmark 57.8% #195 of 216 Official leaderboard

Vision, audio and robotics

BenchmarkScoreRankSource
VAREX-Bench 71% #6 of 8 Official leaderboard
MMAU-Pro 35.4% #11 of 12 Official leaderboard
MMAU v05.15.25 51.7% #7 of 7 Official leaderboard

Who measured it?

Independent runs cover all 14 of its benchmarks.

  • An independent run only 14
Official leaderboards scored it on 14.

Sources: Official leaderboards, Epoch AI (CC BY 4.0) and labs’ model cards