Mostik.ai: Building The Bridge Between AI Models, Very Advanced But Will It Work?

Training a bridge so large and small models share hidden states, exciting, but how will you make money?

The-AI-Brief Weekly Newsletter: Catch Up With All AI's Latest Developments

Everything you need to know from a wild week in AI, from new models and major funding to leaks, pricing and the next big moves.

OpenAI Just Dropped GPT-6. Is It Actually Better?

Sol and Luna are much cheaper, but the early results suggest a smaller upgrade than the GPT-6 name might imply.

Grok 4.7: Why Even Release It?

Better coding cannot hide the high reasoning costs and disappointing overall gains. Who is xAI’s latest flagship actually for?

The Best Use Cases for Jev

Where the decision model wins, the age of Jev

GPT-6 Sol, Opus 5.2 and DeepSeek: AI's Next Big Rumours

I was expecting the next round of AI rumours to be another argument about which model would win the coding benchmarks. Instead, the latest…

Microsoft’s Most Interesting AI Model Isn’t an LLM And Everyone Missed It

Microsoft has a 4 billion parameter AI model that can take a single image and turn it into a high resolution 3D asset, complete with…

Gemini Flash 3.8 Arrived And Again, The Use Case Is Hard To Find. Where Is Gemini Pro?

Google’s new Gemini 3.8 Flash is one of those releases that looks better the longer you stare at the benchmark table and worse the longer…

Claude Code 2.0, The Revamp It Needed! And Higher Limits?? No Not Really..

Claude Code has spent most of this summer getting better in exactly the places where coding agents usually become annoying. Separate…

Fable 5.1 This Week? Gemini 4 And GPT 6? What Can We Expect

At the top of the model market, the gap has become almost absurdly small. Artificial Analysis currently scores Claude Opus 5 at 63, Claude…

Write for The AI Brief: We Are Looking for Writers

Hello and welcome to The AI Brief, an independent Medium publication about the models, systems, infrastructure, economics, and decisions…