Google · AI MODEL

Google: Gemma 4 26B A4B (free)

Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind. Despite 25.2B total parameters, only 3.8B activate per token during inference — delivering near-31B quality at...

Input / 1M: $0.000
Output / 1M: $0.000
Cached input / 1M: —

Specifications

Context window: 262,144
Input modalities: image, text, video
Output modalities: text
Last checked: 10/18/2026, 11:00:41 PM

Source

Prices come from the linked provider source and represent list prices before individual discounts.

Open source

Hugging Face

Inference offers

Compare the live Hugging Face routes available for this model.

No routed inference offers are available yet.

History

Price history

A chart appears after at least two price observations.

Date	Input / 1M	Output / 1M	Confidence
9/7/2026, 7:57:41 AM	$0.000	$0.000	100%

Smart alternatives

Cheaper alternatives

No cheaper compatible alternatives are currently recorded.