The open-source Gemma 4 models are released in four sizes — 31B dense, 26B MoE and effective 2B/4B — optimized for performance, low latency and edge devices respectively. Google DeepMind calls them the best open-source models at their size, emphasizing extremely high intelligence density per parameter.
Key Takeaways
The open-source Gemma 4 models are released in four sizes — 31B dense,26B MoE and effective 2B/4B…
low latency and edge devices respectively.
Google DeepMind calls them the best open-source models at their size,emphasizing extremely high…
This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: Gemma 4 officially released with extremely high intelligence density per parameter 👇. April 3, 2026. https://new.nat6.net/en/news/cmnw1ywov01kyslc3xxzv7090/