blog

News & Updates

Product updates, engineering deep dives, and thoughts on offline AI.

ProductOct 5, 20266 min

Introducing StrataFlow

Run very large Mixture-of-Experts models on CPU, no GPU required. StrataFlow streams experts from disk per token, so a model far bigger than your RAM runs inside a bounded memory budget.

ProductMay 28, 20265 min

Introducing Coaade Code 2.0

Our AI coding engine is here. Highly compressed, fully offline, built specifically for code.

EngineeringMay 20, 20267 min

Why Local-First AI Matters for Enterprise

Cloud AI tools are powerful but come with trade-offs: data leaks, latency, token limits, vendor lock-in. Here's why enterprises are going local-first.

ProductMay 15, 20264 min

The Coaade CLI is Ready

Run AI-powered code generation from your terminal. Ships with the Desktop App.

EngineeringMay 10, 202610 min

How We Compress AI Models 10x Without Losing Quality

A deep dive into our compression techniques that allow Code 2.0 to run on any modern laptop at 95%+ quality.

ProductMay 5, 20266 min

Building the Coaade Desktop App

The Coaade App is your central hub — model manager, product installer, and gateway to local-first AI.

More posts coming soon. Follow us on LinkedIn for updates.