Gemma 4

3 rows in The Deploy Log from 2 companies name Gemma 4: Hugging Face and Google DeepMind. Newest first, each linking the publisher's own page.

Gemma 4 voice AI

Hugging Face and Cerebras demonstrate a real-time speech-to-speech pipeline pairing Gemma 4 31B on Cerebras with Nvidia's Parakeet for speech recognition and Alibaba's Qwen3TTS for text-to-speech, already powering more than 9,000 Reachy Mini robots.

Hugging Face | Models and APIs | In Edition 41, Saturday 4 Jul 2026

SOURCE Hugging Face, Hugging Face and Cerebras bring Gemma 4 to real-time voice AI

Gemma 4 12B

Gemma 4 12B is an encoder-free multimodal model that runs locally with 16GB of VRAM or unified memory, and Gemma 4 models have crossed 150 million downloads.

Google DeepMind | Models and APIs | In Edition 38, Saturday 13 Jun 2026

SOURCE Google DeepMind, Introducing Gemma 4 12B: a unified, encoder-free multimodal model

Gemma 4

Gemma 4 ships in four sizes: Effective 2B, Effective 4B, 26B MoE, and 31B Dense. The 31B ranks #3 among open models on the Arena AI text leaderboard, and the 26B ranks #6. The 26B MoE activates only 3.8 billion parameters during inference.

Google DeepMind | In Edition 28, Saturday 4 Apr 2026

SOURCE Google DeepMind, Gemma 4: Byte for byte, the most capable open models

The edition every row came from arrives in your inbox, free.

Join free