🤖 October Meetup: Running Local LLMs
📅 Monday, October 12, 2026 (rescheduled from Wednesday, October 7)
🕖 7:00–10:00 PM
📍 Kent/Thomas Room @ Unitarian Universalist Church
274 Pleasant Street, Concord, NH
View on map
✨ What to Expect
Missed September? Check out the September 2026 Meetup Recap → — Sam on systems thinking for AI, eight new LoRa nodes on the map, and a new self-hosted home for CRM’s documents.
September looked at how to think about AI. October gets hands-on. Alex presents Running Local LLMs — a practical guide to running models on hardware you own, with no cloud provider in the loop.
Agenda
- 7:00–7:10 — Arrivals & Check-In Go-Around
- 7:10–7:15 — Agenda Review
- 7:15–8:00 — Running Local LLMs — Community Assembly
- 8:00–8:10 — Break
- 8:10–9:10 — Members’ Council
- 9:10–9:15 — Transition
- 9:15–9:45 — Working Group Breakout
- 9:45–10:00 — Closing Circle
🤖 Community Assembly: Running Local LLMs
Presented by Alex
Local LLMs are much more accessible than you would think. They give you a way to take control of your privacy, and they let you AI-assist tasks that the cloud providers will outright refuse. Depending on what you want to do, getting started may be as simple as installing a program on your laptop.
Alex walks through the whole path, from the concepts to a working setup:
- What an LLM is — the high-level concept, without the hype
- What a
.gguf,.safetensor, or.mmprojfile is — and where to get them - Running LM Studio and Ollama — the two easiest on-ramps
- “My hardware is X — what models can I run?” — matching models to the machine you already own
- Tool calls, MCP servers, and RAG — Model Context Protocol and Retrieval-Augmented Generation, and what they unlock
- Uncensored models — why they are a genuinely useful tool, not just a novelty
- OpenRouter and harnesses — how to use OpenRouter, plus a look at Pi, Off Grid AI, LocalMind, and friends
Bring your laptop and your hardware questions. This one is built for following along.
🤝 Members’ Council
Alex will report on filebrowser.capitalmesh.net, CRM’s new self-hosted document server, and the read-only guest access now being set up.
Beyond that: working group reports from LoRa and Reticulum, any other proposals members want to bring forward, and open floor discussion. The LoRa group is still looking for host locations for four Rattlesnake nodes — if you have a site in mind, bring it. To add something to the agenda, email events@capitalmesh.net ahead of time, or raise it during the agenda review.
🛠️ Working Group Breakout & Social Time
After the Members’ Council, we break into open working group time — LoRa, Reticulum, node-host outreach, or just hang out and talk shop.
🧭 RSVP + Directions
The Kent/Thomas Room at the UU Church.
Space is limited — please RSVP by emailing events@capitalmesh.net.
Enter through the East entrance; we’re in the first room on the right.
📅 Add to Your Calendar
Bring a laptop. Leave with your own AI running on it.