Gemma 4 26B A4B
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...
The figures
- Made by
- Model ID
- google/gemma-4-26b-a4b-it
- Input
- $0.068 per million tokens
- Output
- $0.225 per million tokens
- Context window
- 262K tokens
- Longest answer
- 236K tokens
- Takes in
- image, text, video
- Gives back
- text
- Reasoning
- Yes
- Tool calling
- Yes
- Reads images
- Yes
- Open weights
- Yes
- Released
- 3 Apr 2026
Read on 3 Oct 2026.