New models, tools and announcements every day, always with a source link. This digest is assembled by an AI agent following my rules, and it's the same briefing I use myself to keep up with AI without burning time.
Meta has released Glimmer, an open-weight model that anyone can download and run on their own hardware. It contrasts with the more powerful Muse Spark, which the company keeps behind its own APIs. The launch came with a letter from Mark Zuckerberg arguing that AI should be “for everyone”.
Google has released Gemini 3.7 Flash, and the fresh llm-gemini 0.33 plugin already supports it, alongside gemini-3.6-flash, gemini-3.5-flash-lite and two embedding models. Thanks to compatibility with LLM 0.32, you can now see the model’s reasoning traces right in the tool.
OpenAI has launched a preview of Ultrafast, a mode that runs its most powerful model, GPT-5.6 Sol, up to 14x faster. The goal is to court enterprise users for whom response speed is critical.
Google shows how to make private AI practical with homomorphic encryption, a technique that lets you compute directly over encrypted data. The topic resonated strongly with developers.
Suno is unveiling Studio 2.0, pushing its tool closer to a full digital audio workstation (DAW) rather than a bare-bones editor with generative AI. The biggest addition is MIDI support, the company’s most requested feature.
Databricks originally planned to raise $1 billion, but with heavy investor demand it ended up closing $5 billion at a $190 billion valuation. As CEO Ali Ghodsi puts it, AI is expensive.
This section writes, assembles and publishes itself every morning, from source selection to deployment. If you'd like a quiet helper like this in your company, let's start with an initial consultation. The fee is credited toward the project if we work together.