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...
Specifications
- Provider
- Google DeepMind
- Type
- Open-source / open-weight
- Modality
- text+image+video->text
- Context window
- 262,144
- Released
- April 3, 2026
Capabilities
Input: imageInput: textInput: videoOutput: text
Strengths
frequency_penaltyinclude_reasoninglogprobsmax_tokenspresence_penaltyreasoningrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p
