Practical notes on running local AI writing workflows on Apple silicon (16 GB Mac mini): Ollama, llama.cpp, Hugo, and a quiet overnight schedule that leaves Freqtrade dry_run alone.
Affiliate links use /go/<slug>/ short redirects and are disclosed — see Disclosure. Amazon.co.uk tags are placeholders until Associates approval.
Free download: the Quiet Local AI Desk checklist — a printable version of this setup, free on Payhip (my own store; not an affiliate link, no payment required).
Featured buyer guide
Best Mac mini for local LLMs — RAM × model matrix, 16 GB vs 24 GB vs Pro configs, honest limits, and disclosed “See today’s price” CTAs via /go/.
Posts
- Hugo `/go/` Link Patterns for Buyer Posts
- Context Window vs RAM on a 16 GB Apple Silicon Desk
- Avoiding Swap Thrash: Batch Drafting AI Content on a 16 GB Mac mini Without Compromising Freqtrade
- Mac mini vs Mac Studio for Local LLMs
- Ollama on External NVMe: A Safe Symlink Guide for 16 GB Mac Mini
- thunderbolt-dock-vs-usb-hub-mac-mini
- Samsung T7 Shield vs T9 for LLM Weight Stores
- mac-mini-m4-pro-vs-base-for-gguf
- best-keyboard-for-quiet-ai-desk
- best-monitor-for-mac-mini-ai-desk
- Best Mac mini Accessories for a Quiet Local AI Desk (2026)
- mac-mini-m4-16gb-vs-24gb-local-llm
- best-external-ssd-for-ollama-models-uk
- UGREEN vs Satechi Dock for Mac mini M4 Local AI
- When to move from 16 GB to 32 GB for local LLMs
- Mac mini dock vs Samsung T7 Shield for local LLM model stores
- Sizing the Local Desk: A Practical RAM × Model Matrix for 16 GB Mac Minis Running Ollama
- Claims QA Pack — verifying worksheets before any public ship
- Claims sign-off for local-AI stack on Apple silicon
- Deploy rehearsal notes — what a human does on GitHub Pages after sign-off
- Desk close-out after smoke + claims QA — what still blocks a human Pages push
- From dry Hugo build to first GitHub Pages push: what the agent leaves behind
- Human publish gate — pack map before any public push
- Human rewrite before you publish — disclosure, Hugo, and /go/ redirects on a 16 GB desk
- Kokoro TTS disclosure footer on a 16 GB Mac mini
- Pack audio inventory checklist — wav ↔ soft-cite alignment before dry Hugo
- Post-push smoke log — verify /go/ hits without a second deploy
- Pre-push human sign-off card
- Reader Kit QA — empty cites, thin KEEP gates, and local dry builds
- Reader-facing pack map — articles + audio inventory as publish gate
- Smoke-log to claims QA handoff — when a green /go/ hit still needs a human worksheet
- Soft-cite discipline — filling empty [CITE:] without inventing sources
- Voice abstract discipline — ≤120w scripts, Kokoro disclose, no invented claims in audio
- Best Mac mini for local LLMs
- Affiliate allow-list hygiene — links.csv, /go/ redirects, and rejecting invented slugs
- Anti-patterns on a quiet desk — night renders, thermal thrash, and what local AI must never do
- Building a quiet local AI desk on Apple silicon: a practical stack for indie writers
- From desk to pack: assembling a reusable local AI writing kit
- LaunchAgents and the quiet desk schedule — keeping Ollama, overnight video, and Freqtrade in their lanes
- llama.cpp quant choices on 16 GB Apple silicon — when GGUF beats a second UI
- Open WebUI + llama.cpp beside Ollama — when to add a local UI without blowing a 16 GB desk
- Overnight windows and quiet desks — scheduling local inference around Hugo builds and video jobs
- Quiet local AI writing desk on Apple silicon
- Quiet-desk checklist: local AI on 16 GB Apple silicon
- Running local AI on 16 GB Apple silicon: avoiding the second model fight