Google Gemma
Lays out the two branches of Google's open-weight family: Gemma 4 leans on reasoning and local deployment, VaultGemma on differential privacy, each carrying its own licence and compliance calculus.
Lays out the two branches of Google's open-weight family: Gemma 4 leans on reasoning and local deployment, VaultGemma on differential privacy, each carrying its own licence and compliance calculus.
Google's space-based TPU has returned its first AI-generated reply from orbit. On 30 October 2026, Google reported progress on Project Suncatcher: the...
Opera WebAI brings AI models inside the browser. On 27 October 2026, Opera announced in Oslo that it had begun shipping local AI models in its browser...
On October 23, 2026, NVIDIA released Nemotron 3.5 Bolt E4B in an official technical report — an open-source small language model billed as having the ...
NVIDIA and AI4Bharat, the Indian-language lab at IIT Madras, unveiled Nemotron-pretraining-Indic-v1 on October 16, 2026: a pretraining corpus of rough...
Gemma's lightweight model has been put to work fighting scams on phones in Singapore. On October 14, 2026, Google DeepMind announced on its official b...
EmbeddingGemma 2 is an open-weight multimodal embedding model released by Google DeepMind on October 6, 2026 through the Google Developers Blog. Built...
On September 23, 2026, Google announced on its official developers blog that the Antigravity SDK now supports local model workflows, launching with Ge...
Gemma 4 has been released and is positioned as the strongest generation of open models to date. It continues the Gemini 3 homologous research results ...