Back to feed
The Decoder·

Google Deepmind's Gemma 4 12B squeezes multimodal AI onto a laptop with just 16 GB of RAM

Signal
78
Hype
25
In three linesGoogle DeepMind releases Gemma 4 12B, an open-source multimodal model (text, images, audio) running on laptops with 16 GB RAM. Performance nearly matches the 26B model, Apache 2.0 license for commercial use.
Read source
Your take?
DeepMindOpen sourceVisionVoiceBenchmarks

Summary generated by Claude — human-verified