Muse Spark Open: 5 Honest Numbers for a Solo Local Stack
Meta’s 816B open-weight model needs about 304GB even at 2-bit on a Mac. Here is what that means for solo work.
Local LLM
Mistral Large 4: 5 Honest Checks Before Routing Client Work
A trillion-parameter open model with a European legal home. What I’d verify before trusting it with client drafts.
AstaBrief 8B: 4 Smart Signals for Cited Client Research
An open 8B model tuned to cite carefully: four checks before trusting it with client research briefs.
GGUF Quants: 3 Honest Speed Numbers for a Local AI Stack
Hugging Face shipped native GGUF support on September 22. What 40 tokens per second means for a one-person local stack.
Mac Studio M5 Ultra: 512GB Makes a Solid Local LLM Case
The new Mac Studio holds 512GB while the newer M6 chip caps at 32GB. Here is what that flips for local AI.
Muse Glimmer: 4 Honest Signals From Meta’s Local AI Bet
Meta’s new open-weight model is small enough to actually run on my Mac. Here’s why that changes the verdict.
Kimi K3 Open Weights: 5 Honest Signals From the Sidelines
The largest open-weight model yet landed July 27. Here’s why my local LLM setup isn’t switching.
Local AI Privacy: 4 Honest Rules for Client Data
After two months running a local model, here is exactly which client work I keep off the cloud and why.
GLM 5.2 Local LLM Buzz: 3 Honest Numbers From Hacker News
A Show HN post about running GLM 5.2 on a slow machine went viral. Here’s the specs reality check.
Nous Hermes 4: 5 Honest Tests From My First Local Week
Running Nous Hermes 4 locally for client work. Five tests, three pleasant surprises, and the privacy trap I almost stepped into.
Ollama for Solo Consultants: 5 Honest Wins From a Local Setup
Ollama runs my local LLM stack on M-series Mac. Five wins from a solo consultant’s first month — and the workflow it quietly broke.