Find your people.
See how they work.
A look inside the tools, agents and everyday workflows of people building with AI.
Top 20 setups for “local models”
Gareth Price
@gareth…for a rapidly growing data analytics company, using frontier models and a home workstation for local-model inference. The same agent that writes code also administers the machine it runs on. Agents & harnesses Claude Code (current): My…
1,613 viewsNathan Brake
@njbrake…which is OpenAI- and Anthropic-compatible. That lets me put local models and hosted models behind one endpoint and switch per task. Local models do the bulk work. Hosted models do the things local models are bad at: long context,…
396 viewsStevey Brown
@steveybrown…reached over SSH, Tailscale and Moshi. Ollama is installed for local models. Subscriptions & costs Claude Max plan : GBP 200/month Codex Pro 20x : GBP 200/month Wispr Flow Skills Grouped by where they come from, so you can grab the same…
1,819 viewsAlex Martinez
@alexmartinezm…engineer who plans with one agent, implements with open-source models via Pi, reviews with frontier models, and keeps a Basic Memory knowledge base as the source of truth. Alex's setup What I build A B2B SaaS product: .NET backend,…
94 viewsknoopx
@knoopx…enforced via hooks. The whole stack is built to run exclusively on local models: lightweight and lean context. For simple tasks, I start by asking questions about the task I want to achieve, which forces the agent to read and understand…
461 viewso-f
@o-f…An Omarchy desktop (Arch + Hyprland) on a ThinkPad, plus a small local agent fleet. AI helps with desktop questions, a knowledge vault, and one job at a time — I still say yes before anything writes. Agents & harnesses opencode-zen…
242 viewsbluehotdog
@bluehotdog…My editor/IDE alongside the AI tools. LM Studio (current): The local-model side of my setup. Skills & agent instructions Ponytail (active): Pushes the agent toward the smallest correct change: understand the code first, reuse what exists,…
568 viewsKord Campbell
@kord-campbell-kordless…most; I switch harness to fit the job. Open-weight models (current): local models through Ollama — GLM, Qwen, Gemma — via OpenCode and omp, alongside Grok, Claude, Codex and Antigravity. jev (current): TypeSafe.ai's model, used as a…
674 viewsioloro
@ioloro…and file it to Linear, where my agent can pick it up. I use local models for cheap, private per-item judgment, and reserve cloud GPUs for the heavy training step. Before trusting a model's output at scale, I run an adversarial second-look…
345 viewstftio
@tftio…A primary harness that also gives me access to open-weight models through OpenRouter. Clanker (current): My launcher selects context, domain, model and composed instructions; an unresolved context is an error rather than a guess. Models &…
362 viewsJoe Meyer
@jemeyer…performance kernels and data-processing challenges. It queries local models for optimization hypotheses, mutates candidate implementations, and benchmarks them against test suites measuring throughput, latency, and memory. When a candidate…
80 viewsTaylor Houston
@tttaylorrr…can grep it, which beats any vector store for this size of project. Models redline (Overclock) — my default now. One model id that routes to whichever model in the pool is best suited, 260K minimum context. I use it for everything…
223 viewsWes Sander
@ucsandman…edits, driven from Claude Code. OpenClaw (current): Second local agent; a /team skill fans one task out to Claude Code and OpenClaw together. Gemini CLI, agy, grok CLI (installed): Registered as handoff targets for the limit-rotation…
1,626 viewsDru Ibarra
@druwerd…through a change and collect evidence. Docker (current): Runs the local service stack for end-to-end checks. Slack (current): Where teammates post PRs for review, and where the agent picks them up. Models Opus with 1M context (configured):…
902 viewstwKrash
@twkrashtwKrash @twkrash Local-first Pi extension work: an agent harness with delegating, verifying and observing extensions, PR-per-slice review, and evidence-gated releases. I build and maintain Pi extensions — local-first tooling for the Pi…
292 viewsInterestingSoup
@interesting-soupInterestingSoup @interesting-soup Built CRCMZ — a private platform for a 7-person gaming squad: PSN clip pipeline, WhatsApp AI bot, watch party, voice huddle, monthly montages, soundboard, and giveaways. All self-hosted. Claude Code is my…
241 viewsKuyawa
@kuyawa…Agents & harnesses DeepSeek Harness clone in Swift Apps Everything Models deepseek-flash Terminal MacOS Terminal Machines Mac Mini + 60" TV Subscriptions & costs Less than $10/mo Skills All MCP connections None, all tools available in the…
561 viewsJoshua
@joshua-pham…inspect evidence, use tools, and return reviewable results across local or remote environments. Obsidian : This is my durable, human-readable source of truth for notes, research, strategy, plans, and spatial thinking. I treat it as…
656 viewsKW Na
@member-2c840fe7KW Na @member-2c840fe7 Kotlin/Spring Boot 백엔드 개발자. IntelliJ + Claude Code, 레포에 규칙·리뷰 에이전트·훅을 넣어 컨벤션을 하네스로 강제한다. 내가 하는 일 Kotlin · Spring Boot 기반 멀티모듈 백엔드 모노레포(API, 인증 BFF, 배치)를 개발한다. AI는 설계 논의, 구현, 마이그레이션 작성, 커밋 전 리뷰에 쓰고, 최종 판단과 코드 확인은 내가…
19 viewsSrijan Shukla
@srijanshukla18…mention the user journeys for the product, then I make the frontier models develop a High Level Design and Mid Level Design - which I read and review myself. Once I ask for changes - then I leave the code to a pareto frontier model like…
645 viewsCurious about someone’s AI setup? Ask them to share it.
Add their @mentions in X.