Google

gemma-3-4b-it

SIZE4.3B

21Quality

Vision

Downloads1.1M/mo

Likes1.6K

Vision-language model from Google, 4.3B parameters.

Open weightsGemma license501/1044

Drag to turn

GoogleTextPromo

gemma-3-4b-it

Vision-language model from Google, 4.3B parameters.

google/gemma-3-4b-it

Type: Text
Quality: 21 · Top 79% on the Epoch Capabilities Index
Parameters: 4.3B (4,300,079,472)
Downloads: 1.1M in the last 30 days
Likes: 1.6K
Access to weights: Accept the license on Hugging Face first
On Hugging Face since: Feb 20, 2025
Takes in: Text, Image
Puts out: Text
Access: Open weights
License: Gemma license
Rarity: Promo
Set number: 501/1044

Open on Hugging Face Quality 21 Where to run it 12 Evolution 2/4 Race it 26 tok/s Meet its creature Battle GLM-4.7-Flash

License read from the model’s Hugging Face card. Stats from Google’s official Hugging Face account, refreshed daily.

How good is it?

Public leaderboard scores for gemma-3-4b-it, refreshed daily from Epoch AI and LMArena.

Top 79% of the 277 models on the Epoch Capabilities Index: #218.

Quality is the share of a leaderboard’s models this one beats, from 0 to 100. Chat models are rated on Epoch AI’s Capabilities Index when it lists them, otherwise on LMArena; image and video models on LMArena’s image and video arenas.

Leaderboards

Epoch Capabilities Index#218 of 277
One number fitted from dozens of hard benchmarks, from science questions to agentic coding.116.0 index
Listed as Gemma 3 4B
LMArena Text#255 of 416
People compare two anonymous answers and vote for the better one.1,291 rating
LMArena Text · Coding#301 of 411
The text arena’s coding prompts only.1,230 rating
LMArena Text · Math#285 of 400
The text arena’s math prompts only.1,239 rating
LMArena Text · Creative writing#247 of 414
The text arena’s creative writing prompts only.1,270 rating

Benchmarks Epoch AI runs

Scores from Epoch AI and LMArena, licensed CC BY 4.0. Boards list models under their own names, so a close match is shown with the name it was scored under. Credits

Where to run it

10 providers sell gemma-3-4b-it through an API, and 2 Hugging Face hosts serve it. Prices are per million tokens, cheapest first.

CheapestKilo Gateway$0.04 in · $0.05 out
Free to try1 providerNvidia
Fastest on Hugging FaceDeepInfra26 tokens/s

API providers

Provider	Input	Output	Context
NvidiaFree`google/gemma-3-4b-it`	Free	Free	131K
Kilo GatewayCheapest paid`google/gemma-3-4b-it`	$0.04	$0.05	131K
Amazon Bedrock`google.gemma-3-4b-it`	$0.04	$0.08	131K
Eden AI`amazon/google.gemma-3-4b-it`	$0.04	$0.08	128K
Merge Gateway`google/gemma-3-4b-it`	$0.04	$0.08	128K
Deep Infra`google/gemma-3-4b-it`	$0.05	$0.10	131K
OpenRouter`google/gemma-3-4b-it`	$0.05	$0.10	131K
AI/ML API`google/gemma-3-4b-it`	$0.069	$0.14	131K
Pioneer`google/gemma-3-4b-pt`	$0.15	$0.15	33K
NanoGPT`unsloth/gemma-3-4b-it`	$0.20	$0.20	128K

Hugging Face Inference Providers

Host	Input	Output	Context	Speed	First token
DeepInfraFastestTools	$0.05	$0.10	131K	26 tok/s	0.4s
Featherless	—	—	—	—	—

From models.dev and Hugging Face, checked daily. Resellers’ listings can lag behind their price pages, so confirm with the provider before you commit.

gemma-3-4b-it

gemma-3-4b-it

How good is it?

Leaderboards

Benchmarks Epoch AI runs

Where to run it

API providers

Hugging Face Inference Providers

More from Google

Gemini 4 Argon

embeddinggemma-2

Gemini 3.8 Flash

timesfm-3.0-pytorch

Gemini 3.7 Flash

Gemini 3.6 Flash

Gemini 3.5 Flash Lite

Nano Banana 2 Lite