●NANOLITE — Nano Banana 2 Lite is here: Google's fastest and most cost-efficient Gemini Image model, made for running lightweight image generation cheaply●OMNIFLASH — Gemini Omni Flash is in public preview, a natively multimodal model that lets enterprises and developers build custom, dynamic video workflows●AGENTS — Managed Agents expand with background: true for async server-side runs and polling, remote MCP server integration, and refreshing credentials across interactions●MEMORY — The Memory Bank IngestEvents API is generally available, decoupling event ingestion from memory generation so you can stream content continuously●THROUGHPUT — Provisioned Throughput now lets you submit up to seven pending orders for the same model and region●DEPRECATE — Image generation models shut down on August 17, and the Grok 4.1 family on the Gemini Enterprise Agent Platform on August 20●NANOLITE — Nano Banana 2 Lite is here: Google's fastest and most cost-efficient Gemini Image model, made for running lightweight image generation cheaply●OMNIFLASH — Gemini Omni Flash is in public preview, a natively multimodal model that lets enterprises and developers build custom, dynamic video workflows●AGENTS — Managed Agents expand with background: true for async server-side runs and polling, remote MCP server integration, and refreshing credentials across interactions●MEMORY — The Memory Bank IngestEvents API is generally available, decoupling event ingestion from memory generation so you can stream content continuously●THROUGHPUT — Provisioned Throughput now lets you submit up to seven pending orders for the same model and region●DEPRECATE — Image generation models shut down on August 17, and the Grok 4.1 family on the Gemini Enterprise Agent Platform on August 20

TAG

Edge AI

2 articles

← Back to all tags

Gemma 4² MoE¹ Dense¹ Quantization¹ Performance¹ Local LLM¹ AI Development¹ Cost Optimization¹

⬡ Gemini Advanced/2026-04-12Advanced

Gemma 4 MoE vs Dense: Architecture Selection and Performance Optimization Guide

Deep dive into Gemma 4's 26B MoE vs 31B Dense: measured benchmarks, use-case selection criteria, quantization strategies, and deployment configurations from edge to cloud.

◎ Updates/2026-04-11Intermediate

Run Gemma 4 Locally: Building Zero-Cost AI Apps

A complete guide to running Gemma 4 locally and building AI apps with zero API costs. Covers setup on MacBook and Raspberry Pi, production architecture design, and leveraging the 256K context window.