GoogleReleased on May 23, 2026

Gemma 4 12B: API Pricing, Context Window & Benchmarks

Gemma 4 12B is a language model from Google, released in May 2026, with multimodal input.

Gemma 4 12B is Google DeepMind's encoder-free multimodal instruction-tuned model with 11.95 billion parameters and a 256K context window. It supports text, image, audio, and video inputs with text output, projecting image patches and audio

Gemma 4 12B benchmarks

Rankings

Quality Tracker

Gemma 4 12B Performance Across Datasets

Scores sourced from the model's scorecard, paper, or official blog posts

LLM Stats Logollm-stats.com - Fri Jul 24 2026
Notice missing or incorrect data?

Gemma 4 12B model size

Gemma 4 12B has 12.0 billion parameters. See how it compares to other models in the same parameter range.

Parameters
12.0B
Medium (10–30B)
12.0B
1B7B70B405B

Gemma 4 12B API

Available from the model provider

Gemma 4 12B has an official provider API. It is not currently routed through the LLM Stats gateway.

Read the official API documentation

Gemma 4 12B latency

Gemma 4 12B time to first token, sustained output throughput, and failed-request rate from live API traffic over the trailing 7 days.

Gemma 4 12B examples

Recent arena outputs from Gemma 4 12B, picked from the highest-ranked matchups.

Gemma 4 12B license

Gemma 4 12B is released under the Apache 2.0 license, which permits commercial use, has 12.0B parameters, has a knowledge cutoff of January 2025.

License
Apache 2.0
Commercial use allowed
Parameters
12.0B
Knowledge cutoff
January 2025

Apache License 2.0 - allows commercial use

Gemma 4 12B resources

Official sources for Gemma 4 12B: api documentation.

Gemma 4 12B vs other models

The most-compared alternatives to Gemma 4 12B are Claude 3.5 Sonnet, o1-pro, Qwen3 VL 235B A22B Instruct. Open any pair side-by-side for benchmarks, pricing, context, and latency.

Models like Gemma 4 12B

Models ranked just above and below Gemma 4 12B by LLM Stats score.

 

Claude 3.5 Sonnet

Score pending
DetailsCompare
 

o1-pro

Score pending
DetailsCompare
 

Qwen3 VL 235B A22B Instruct

Score pending
DetailsCompare
 

Qwen3 VL 32B Thinking

Score pending
DetailsCompare
 

Sarvam-105B

Score pending
DetailsCompare
 

o1

Score pending
DetailsCompare

FAQ

Common questions about Gemma 4 12B.

When was Gemma 4 12B released?

Gemma 4 12B was released on May 23, 2026 by Google. This is the official Gemma 4 12B release date tracked on LLM Stats.

Is Gemma 4 12B available via API?

Yes, Gemma 4 12B is available via API. See the official documentation for authentication and endpoint details.

How big is Gemma 4 12B?

Gemma 4 12B has 12.0 billion parameters. It ships as an open-weight model, so you can download and run it on your own hardware.

Who created Gemma 4 12B?

Gemma 4 12B was created by Google.

What is the license for Gemma 4 12B?

Gemma 4 12B is released under the Apache 2.0 license. This is an open-source / open-weight license that permits self-hosting.

What is the knowledge cutoff date for Gemma 4 12B?

Gemma 4 12B has a knowledge cutoff of January 2025, meaning it was trained on data up to that point and may not know about events after it.

Is Gemma 4 12B multimodal?

Yes, Gemma 4 12B is multimodal and can accept both text and images as input.

Where is the Gemma 4 12B paper or technical report?

Gemma 4 12B has a paper or technical report available at https://huggingface.co/google/gemma-4-12B-it. Use that source for architecture, training, release and evaluation details.

What models should I compare Gemma 4 12B against?

Common Gemma 4 12B comparisons include Gemma 4 12B vs Claude 3.5 Sonnet, Gemma 4 12B vs o1-pro, Gemma 4 12B vs Qwen3 VL 235B A22B Instruct. Compare them side by side for benchmark scores, pricing, context window, latency and API availability.