Trending News

4 trending stories

“#Gemma”×All Time×

Google releases Gemma 4 12B, an open-source model capable of local multimodal inference

Google's new Gemma 4 12B model allows for local audio and video analysis on 16GB laptops, potentially reducing infrastructure costs for AI-focused startups.

#Startups#AI#OpenSource

Jun 23, 20261m readAI

VentureBeat

Google Gemma models exhibit reasoning failure when processing resource-constrained scenarios

Trending

💼 Business & Finance

Google Gemma models exhibit reasoning failure when processing resource-constrained scenarios

New evidence suggests Google’s Gemma models struggle with basic resource-tracking tasks, potentially limiting their utility in multi-step planning and real-world reasoning applications.

#StartupsEntrepreneurship#Gemma#LLM

Developer achieves 124 tokens/sec with Gemma-4 26B on CPU-only hardware

A new optimization technique for the Gemma-4 26B model achieves 124 tokens/sec on CPU-only setups by prioritizing output head compression over expert compression.

#StartupsEntrepreneurship#Gemma#LLMOptimization

Cerebras adds multimodal support for Gemma 4 models on WSE-3 hardware

Cerebras brings multimodal support to Gemma 4, targeting high-speed inference. We analyze the claims versus the need for independent performance validation in enterprise settings.

#StartupsEntrepreneurship#Gemma#Cerebras

Jul 3, 20261m readAI

Hacker News: Newest

All caught up · 4 stories