Kicking off a series on replacing a Samsung SmartThings hub with a self-hosted Home Assistant box. Episode 1: picking the hardware, fighting a Lenovo BIOS setting hidden behind a decoy, and the increasingly absurd path to actually getting an OS onto the machine when there wasn’t a single working USB drive anywhere in reach.
Three models independently audited the same pinned code in isolated workspaces, blind to each other. Two produced reports. Opus 5 found the bug that actually mattered. Fable 5 was blocked twice, at the same spot.
Meta had the one asset nobody else in the West could match — frontier-scale open weights, right as sovereign AI became a real enterprise buying criterion. Instead of owning that lane, it built Muse Code to fight Anthropic, OpenAI, and now xAI for a shrinking slice of the most contested market in AI.
My gaming rig’s Windows partition used to cost me 15-45 minutes of manual clicking every time I booted into it. One afternoon: a Coder workspace talking to it over Tailscale and SSH, an inventory script that catalogs every update channel on the machine, three Scheduled Tasks that actually do the work, a toast notification bug that took longer to fix than the automation itself, and a Next.js dashboard on Vercel to see it all without ever touching the machine.
My Windows gaming rig, hostname Stormtrooper, had a case fan installed backwards. Turning it around dropped GPU temps 16-18°C. Then I went looking for a CPU undervolt and found out the board doesn’t have one. Then I tried to make idle quieter and accidentally cooked the CPU 18°C hotter across every phase, including the ones with no load at all. Four thermal test runs, one Core Ultra 9 285K, and a lot of honest data.
A weekly homelab housekeeping pass (Ollama out, Home Assistant current, RustDesk re-pulled), a quote-marks bug that turned out to be a content pipeline gap rather than a font problem, and a full audit of this blog’s own “By the Numbers” habit that found it wasn’t as consistent as assumed.
My wife saw OpenAI’s new Codex Micro and thought she could skip buying one — she already has a Stream Deck. That sent me down a rabbit hole: what the device actually does, how it compares to the Elgato hardware it’s built on top of, and whether a Linux homelab version is a real project or just another case of reinventing a wheel. A survey first, a build plan second, no results yet.
A challenge to a coding agent: prove whether months of homegrown test harnesses and bakeoff scripts were reinventing wheels that already exist. A survey of what hobbyists, techtubers, and the AI benchmarking industry already publish, and an honest verdict on this blog’s own tooling habits.
The Linux Foundation just launched the Tokenomics Foundation to extend FOCUS — the FinOps billing spec that normalized cloud cost data — into token-based AI spend. It’s the cloud-native parallel playing out again, except this time every knowledge worker who touches an agent is about to become a cost center. Here’s what FOCUS 1.4 actually standardizes, what it still can’t see on self-hosted infrastructure, and why we’re going to try to run it on a homelab anyway.
I put Qwen 3.6 27B, Qwen 3.6 35B-A3B, Qwythos-9B, GLM-4.7-Flash, and Nemotron-3-Nano through the same real coding task on my homelab RTX 5090. Along the way I had to live-patch two separate llama.cpp bugs — and even after fixing them, I couldn’t fully prove one model’s failure wasn’t the harness’s fault.
Back in Q1 I read a headline about Google cutting AI memory use 6x and filed it under “watch and revisit.” Four months later, Google still hasn’t shipped official code, but a whole ecosystem of forks has, llama.cpp has an open PR, and at least one compatibility gotcha lands squarely on our daily driver. Here’s the honest state of TurboQuant heading into Q3, and the test I’d actually run against it.
A routine “update Coder” request turned into a full model bakeoff: fixing misconfigured thinking params so Sonnet 5 would stop calling itself 4.5, discovering Playwright MCP can’t be wired into Coder Agents at all, and watching three frontier models independently pause at the exact same step before finishing. Sonnet 5 won on score and on cost, by a mile.
Mid-research on the LocalAI bakeoff, I got asked a blunt question: why not just run Coder Agents as the homelab supervisor too? The answer split my original coding-vs-general-purpose hypothesis into a sharper dichotomy — ephemeral/invoke-driven/git-centric vs. persistent/event-driven/tool-centric — and pointed straight at Turnstone as the next thing worth a real bakeoff.
A gloomy Cape Cod afternoon post-July 4th turns into a deep dive on ComfyUI, Lemonade Server, and LocalAI — plus llama-benchy and AMD’s AI Playbooks — and the case for a bakeoff against our hand-tuned llama.cpp stack.
GLM is suddenly everywhere in developer conversations. Before we run the bakeoff, we need to answer two questions: what is GLM, and is it suitable for a single RTX 5090 homelab?
A full-tower AI homelab with a 420mm AIO gets rebuilt into an SFF open-frame case with custom hardline water cooling. Two 240mm slim radiators, a single shared loop, 580W peak heat load, and 2,726 sensor readings that prove whether the tradeoff was worth it.
I gave five local LLMs and one frontier cloud model the same coding task on my homelab: build a tag manager for the blog’s admin panel. Only two shipped anything. Here’s what happened.
Two Discord bots, one 14B model, five fitness-tracker tasks. Both agents failed on the first try. Getting them working required debugging context overflow, silent tool parameter drops, and a chat template flag that changes everything. The results reveal as much about the state of local AI agents as they do about which framework won.
Coder 2.34 shipped User Secrets — per-user credential storage that injects into every workspace automatically. We upgraded, audited 29 secrets across four projects, and found exactly two that belonged there. Here’s how we decided, how we migrated, and what we cleaned up along the way.
A full walkthrough of setting up Wake on LAN on a Linux homelab and wiring it into Google Home via SmartThings — including every dead end, expired link, and wrong interface name along the way.
Two weeks of using Qwen3.5-35B as my daily AI assistant — the Jinja template fix that made it work, the thermal spam incident that almost ended the experiment, and the session-context gap that makes it feel like a junior dev every morning. Plus: what’s next with Qwen 3.6.
I built a Model Context Protocol server into the fitness tracker I vibe coded a year ago, wired it through Vercel and Coder workspaces, and ended the afternoon asking my Discord bot what my last workout was. Here’s the build, the wrong turn into Coder’s AI Bridge, the workaround, and how the same endpoint now serves Claude Desktop, Codex, Coder Agents, and OpenClaw.
From curl to working Discord bot in one afternoon — with a local LLM on the RTX 5090. Every gotcha, every config mistake, and the one setting that silently ate every server channel reply for hours.
DeepSeek V4-Pro, V4-Flash, and Zyphra ZAYA1 are three of the most exciting new models in local AI. None of them run on our RTX 5090 homelab — for completely different reasons. Here’s the research, the math, and what it means for anyone building a local inference rig.
A second user joined the homelab Coder instance and couldn’t push to GitHub. What looked like a missing config turned into five chained problems, a domain migration aftershock, an agent-debugging-an-agent meta-moment, and the discovery that the same credential helper bug had been “fixed” four times in ten days — and never actually deployed.
We ripped out Ollama, migrated to llama.cpp, and benchmarked five local models across 12 tasks on an RTX 5090. The results surprised us — and the winner wasn’t who we expected.
I asked an AI agent to turn off my RGB lights on Linux. 85 terminal commands, 35 failures, 4 hangs, 2 dead download links, one wrong build system, and the GPU is still glowing. This is the post.
Gemma 4 failed to build a single feature in our last test. This time we diagnosed the problem, switched from Ollama to llama.cpp, tuned the inference settings, and Gemma shipped a working search feature to production. Then Opus reviewed the code and made it better. Here’s what we learned about making local models actually work.
Four bugs that were silently breaking things for days: a deploy that only crashes on new images, a shell guard that eats your auth tokens, a publish date frozen at draft creation, and a homelab with no emergency remote access. Plus: capacity planning for when you’re running AI workspaces on a single machine.
We pitted Gemma 4 against Opus 4.6 on a real feature build for vibescoder.dev. Gemma is the fastest model in our benchmark. It also couldn’t finish the job. Here’s what happened when we stopped testing toy apps and started building production code.
We added Google’s Gemma 4 and Moonshot’s 1-trillion-parameter Kimi K2 to the local model benchmark. Five out of six models scored perfect. Gemma 4 is the new speed king. And yes, we ran a 579 GB model off an NVMe drive — at 0.6 tokens per second.
While waiting for massive open source models to download, I tackled the homelab backlog: custom domain for my Coder instance via Cloudflare Tunnel, security hardening (with a gotcha that could kill your AI search visibility), and wiring up MCP servers to give agents superpowers.
We gave six LLM models the exact same coding prompt and measured everything: speed, tokens, and whether the code actually works. Three models scored perfect. Two built the wrong kind of app. One ran out of tokens mid-line.
Installing Ollama, pulling five purpose-built models, wiring local inference into Coder Agents, and running agentic coding on an RTX 5090 workstation. 44 GB of models, zero cloud API calls, fully self-hosted.