Skip to content
← Models
Open weightsGoogle/10 trustVerified
Gemma 3 logo

Gemma 3

Google's open-weight multimodal family, 1B to 27B, with a single-GPU friendly footprint.

Visit official page →

About

Gemma 3 is Google's third-generation open-weight model family, released March 12, 2025 in four sizes: 1B, 4B, 12B, and 27B parameters. It natively handles text and images, supports over 140 languages, and offers a 128K-token context window on the larger sizes (32K on the 1B model). The 27B variant was designed to fit on a single GPU or TPU, making it one of the most capable models runnable on modest hardware.

An honest caveat for the catalog: Gemma's weights are downloadable and free to use, but they ship under Google's Gemma Terms of Use — a custom license that is not OSI-approved open source and carries use restrictions. It is open-weight, not open-source in the strict sense.

Gemma 3 is positioned as the open counterweight to Google's closed Gemini models, sharing much of the same research lineage while letting developers run, fine-tune, and deploy the weights themselves.

Specs

Context window

Pricing

Free quota

API access

Weights

Open weights

Released

March 12, 2025

Benchmarks

Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.

Use cases

chatbotsmultimodal-aion-device-aifine-tuning

Trust score — 8/10

/10 trustVerified

Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 73; Trending score 82/100 from researcher assessment.

More in the gemma family

Sources