Gemma 3 1B

Gemma 3 1B.

Google DeepMind Released 12 Mar 2025

Benchmarks with a score

Ranked first, out of five or more models

In the top three

Google DeepMind Mar 2025

Open weights

No Epoch Capabilities Index score yet

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.

Benchmark	Score	Rank	Source
Vectara Hallucination Leaderboard (HHEM)	5.30%	#53 of 148	Official leaderboard
IFEval	80.2%	#98 of 126	Model card
BIG-Bench Hard	39.1%	#66 of 71	Model card
SIQA	48.9%	#21 of 22	Model card
PIQA	73.8%	#32 of 33	Model card
TriviaQA	39.8%	#38 of 39	Model card
WinoGrande	58.2%	#41 of 42	Model card
ARC-Challenge	38.4%	#44 of 45	Model card
HellaSwag	62.3%	#46 of 47	Model card
MMLU-Pro	14.7%	#151 of 151	Model card
MMLU	38.8%	#127 of 127	Model card
SimpleQA	2.20%	#59 of 59	Model card
MGSM	2.04%	#52 of 52	Model card
BoolQ	63.2%	#28 of 28	Model card
ARC-Easy	73%	#23 of 23	Model card
Natural Questions	9.48%	#20 of 20	Model card
Global MMLU Lite	34.2%	#16 of 16	Model card
FACTS Grounding	36.4%	#15 of 15	Model card
BIG-Bench Extra Hard	7.20%	#11 of 11	Model card
ECLeKTic	1.40%	#10 of 10	Model card
Flores	29.5	#8 of 8	Model card
IndicGenBench	41.4	#4 of 4	Model card
XQuAD	43.9	#4 of 4	Model card

Benchmark	Score	Rank	Source
MATH	48%	#67 of 93	Model card
GSM8K	62.8%	#73 of 87	Model card
OTIS Mock AIME 2024-2025	1.11%	#194 of 202	Epoch AI
GPQA Diamond	19.9%	#215 of 219	Epoch AI
HiddenMath	15.8%	#11 of 11	Model card

Benchmark	Score	Rank	Source
HumanEval	41.5%	#84 of 95	Model card
MBPP	35.2%	#62 of 66	Model card
LiveCodeBench	1.90%	#73 of 73	Model card
Bird-SQL	6.40%	#13 of 13	Model card
Natural2Code	56%	#9 of 9	Model card

Benchmark	Score		Rank	Source
Berkeley Function Calling Leaderboard (BFCL)	16.3%		#67 of 67	Official leaderboard
BFCL V4	7.17%		#65 of 65	Official leaderboard

Every result in the longer documents the lab published with it. 33 of them aren't in the launch post.