gemma-4-31B-it
GoogleBenchClub cloud
No cloud result yet
Headline numbers come from BenchClub’s pinned cloud runner.
- Arena score
- 1443
- Among open models
- #15of 148
- Among all on Arena
- #73
- Votes
- 6,130
Measured on Gemma 4 31B, the served version, not a local build. LMArena, CC BY 4.0, published Nov 11, 2026.
Score by category
Top 10 among open models in#6Russian
CategoryScoreOpen rankArena rankVotes
- Chinese1476#23#85521
- Expert1465#15#70501
- Applied math1464#14#64344
- Software & IT1463#18#762.2K
- Math1462#13#53430
- Russian1460#6#42824
- Coding1459#23#871.4K
- Science1456#19#72930
- Hard prompts, English1456#15#741.3K
- Multi-turn1452#11#661.1K
- English1451#15#712.3K
- Medicine & health1450#22#78421
- Hard prompts1447#19#843.6K
- Long prompts1445#13#721.8K
- Spanish1444#20#76253
- Excluding ties1441#13#674.2K
- Legal & government1438#24#96451
- French1434#31#109267
- Instructions1433#12#721.8K
- Polish1431#21#84203
- Writing & literature1431#12#681.4K
- Business & finance1431#18#791.2K
- Non-English1430#14#723.8K
- Creative writing1415#13#76975
- Media & sports1405#16#871.1K
| Model | Params | Score |
|---|---|---|
| 25.8B | 1434 | |
| 31.3B | 1443 |
BenchClub runs
Discussion
- GEMMA4 DE GOOGLE - IA EN LOCAL GRATIS EN TU MOVIL#ia #ai #gemini #bigdata #data #bilbao #gemmamaxmaxdata · Aug 27, 20266.5K likes
- Google releases Gemma 4 open modelsjeffmcjunkin · Apr 2, 20261.8K points
- Gemma 4 12B: A unified, encoder-free multimodal modelrvz · Jun 3, 20261.1K points
- 🤯 ना ChatGPT, ना Gemini! Google का ये Secret AI बिना Internet चलेगा 😱 | Gemma 4 FREE!mybigguide · Oct 24, 20261K likes
- Show HN: Open-source engine running Gemma 4 26B in 2 GB RAM on any M-series Macgitpusher42 · Jul 29, 2026919 points
Built on it
Ways to run it
- GGUF · llama.cpp, Ollama, LM Studiounsloth/gemma-4-31B-it-GGUF152 builds
- MLX · Apple silicon: mlx-lm, LM Studiolmstudio-community/gemma-4-31B-it-MLX-8bit65 builds
- AWQ · vLLM, SGLangQuantTrio/gemma-4-31B-it-AWQ7 builds
- GPTQ · vLLM, SGLangpositron-ai/google_gemma-4-31B-it-ingest-best-gptq7 builds
Most downloaded
More from Google
See more
gemma-4-26B-A4B-it5.1M25.8B

gemma-3-4b-it995.4K4.3B

gemma-3-12b-it409.4K12.2B

gemma-3-27b-it307.6K27.4B
Other 20–40B models
See more
Qwen3.8-27BQwen6.9M27.8B
Ornith-1.0-35BOrnith1.4M35.1B
NVIDIA-Nemotron-3-Nano-30B-A3B-BF16NVIDIA1.3M31.6B
GLM-4.7-FlashZ.ai1.2M31.2B