- IF Eval 0-shot (IT) 51.2
- BIG-Bench Hard few-shot (IT) 26.7
- HellaSwag 0-shot (IT) 37.7
- PIQA 0-shot (IT) 66.2
- ARC-c 0-shot (IT) 28.2
- WinoGrande 0-shot (IT) 52.3
Gemma 3 270M
Gemma 3 270M.
6
Benchmarks with a score
0
Ranked first, out of five or more models
0
In the top three
Google DeepMind Aug 2025 Gemma 3 270M
Open weights
70.9
#404 of 404
Self-Reported Index, from the lab's own numbers. No Epoch Capabilities Index score yet.
- IFEval 51.2% #134 /135
- BIG-Bench Hard 26.7% #73 /73
- HellaSwag 37.7% #53 /53
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
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| IFEval | 51.2% | #134 of 135 | Lab model card | |
| BIG-Bench Hard | 26.7% | #73 of 73 | Lab model card | |
| HellaSwag | 37.7% | #53 of 53 | Lab model card | |
| ARC-Challenge | 28.2% | #49 of 49 | Lab model card | |
| WinoGrande | 52.3% | #47 of 47 | Lab model card | |
| PIQA | 66.2% | #37 of 37 | Lab model card |
Its lab's numbers put it at 70.9, number 404 of 404
The Self-Reported Index fits one score to the 6 capability results Google DeepMind reported for this model, on the scale of the Epoch Capabilities Index. Labs choose which benchmarks to show and how to run them, so read it as the lab's claim, not an independent measurement. The 90% interval runs from 62.1 to 79.8. How we compute it.
- PIQA 66.2
- WinoGrande 52.3
- IFEval 51.2
- HellaSwag 37.7
- ARC-Challenge 28.2
- BIG-Bench Hard 26.7
Its launch post
The benchmarks the lab chose to headline, with the scores it gave.