A "memory computer" for the home: your photos, files and messages on a swappable Memchip, with a local 27B model working on them 24/7. $1,799 at launch, no subscription.

An AI browser built on Chromium that does the work, not just the chat: research, prospecting and ops tasks that run for minutes or hours, logged in as you. Free to start, $20 a month.

A $3,499 box that runs open models at home 24/7, connects to your apps, files and devices, and acts without being asked. No subscription, nothing leaves the house.

Aleph Alpha's open-weight reasoning model, built and trained in Europe: 78B mixture-of-experts with 3.5B active, German and English, up to 1M tokens of context, Apache 2.0.

TypeSafe AI's Jev decides instead of writing: a typed answer with a probability in about 150ms, where an LLM takes seconds.

Fermion Research's open English speech-to-text model: Parakeet-level accuracy in a 164 MB file, 174 times realtime on a MacBook Air, CC-BY-4.0.

A running collection of AI-generated videos worth keeping: parodies, short films and experiments, each hosted here with its credit and a line on why it's in the list.

Dutch cloud service for full Linux VMs that boot in milliseconds, fork while running and sleep when idle: persistent computers for developers and AI coding agents.

French AI inference service on European GPUs: open-weight models behind an OpenAI-compatible API, no data retention, $0.12 per million input tokens and 1 billion free tokens.

An AI testing platform for web, iOS and Android apps: tests are plain-English YAML steps in your repository, and they repair themselves when the UI changes.

Tencent's open-source 1.5B speech model: text to speech, voice cloning, and editing of recorded speech (words, pitch, speed, emotion, noise) from plain-language instructions.

One API key and one bill for 458 AI models from 63 labs, with automatic fallbacks when a provider fails. Model prices pass through, and OpenRouter takes 5.5% on credit purchases.

NVIDIA's open-weight 100M-parameter model that tells who spoke when, for up to 8 speakers, live or recorded. Ranked first on VoiceArena's diarization benchmark at launch.

Anthropic's guidance for getting the most out of each Opus release in Claude and Claude Code, newest first: Opus 5.5 (September 2026) and Opus 4.7 (April 2026).

Dwarkesh Patel's 80-minute interview with OpenAI's Noam Brown: how 10,000-agent swarms work, what maths progress says about RSI, and the alignment problem it raises.

Pi is a minimal, open-source, bring-your-own-key coding agent harness, and the harness alone can cut cost about 2x for the same model.

Nodes, edges, the fake-edge test and the diamond, from a long-form X post, checked against how Claude Code actually runs agents in parallel.

Cloudflare wrote a browser engine in Rust that runs inside Workers V8 isolates. 3-7x less CPU and memory than Chromium for agent tasks.

A private, HIPAA-compliant ChatGPT alternative for healthcare: AI scribe, clinical notes, and document analysis over frontier models with a signed BAA.

How context rot quietly degrades Claude's outputs, how to track it with /context, and why a refresh skill beats /compact for carrying a task into a fresh chat.

Generative media platform with one API for the best image and video models (Flux 2, Nano Banana Pro, Ideogram), wired into NicAI to make sales-deck visuals on demand.

Free, open-source, on-device voice-to-text for macOS with a choice of local models (Parakeet, Whisper, Apple Speech), pitched as a Wispr Flow alternative.

Ben AI's De-Slop skill, a quality gate that checks any AI output before it ships, and how it compares to my own anti-AI-writing guidelines.