When I rebuilt my home server, my goals were:

- To make sure that the #Ollama instance was using the GPU
- Move the Dell 1L PC that #Microsoft EOLed onto a Windows 11 VM
- Add enough disc storage to eliminate the need for the #Drobo

I wasn't expecting to self-host many things. So far, I am self-hosting ...

- Open WebUI
- Uptime Kuma
- Immich
- n8n
- Appwrite
- Jupyter

... using real subdomains connected to 100. Meshnet IP addresses. It means that each web app should work seamlessly, but only from one of my devices. I like that a lot.

New Ollama Release!
Version: v0.8.0
Release Notes:

## What's Changed
* Ollama will now stream responses with tool calls
* Logs will now include better memory estimate debug information when running models in Ollama's engine.

## New Contributors
* @hellotunamayo made their first contribution in https://github.com/ollama/ollama/pull/10790

**Full Changelog**: ...
https://github.com/ollama/ollama/releases/tag/v0.8.0

#Clara - Fully Local #AI Superstack with #Ollama #LLM Chat, Agent Builder & #StableDiffusion

Complete #opensource #AI workspace running 100% offline on your machine
Features #LLM chat with #Ollama models, visual app builder, and #N8N automation integration

So I had some really cool stuff going on with an optional add-on for Chit (my #Ollama LLM frontend for Ollama) that would pull in content from the web for context.

But I've hit up on DuckDuckGo's rate limiting, which poured water on the good times. It was working decently, too.

Any advice on legit ways to get low-volume single-person search engine results? Just need the top 5 URLs for a search query.

There’s Proof! AI Makes You Stupid. https://indubitablyodin.medium.com/theres-proof-ai-makes-you-stupid-55b08486efee

Here's a screen cast of the gemma3 1b model running in Ollama in a LXD container on Ubuntu Touch on my Pixel 3a device.

Your #1 Ethical AI List https://medium.datadriveninvestor.com/your-1-ethical-ai-list-7d306816e6f2

Am Mac verwende ich aktuell #raycast und GPT-4.1 mini um mir eine Bildbeschreibung zu meinen Fotos erstellen zu lassen. Das klappt gar nicht mal so schlecht. Jetzt habe ich mal wieder #ollama angeworfen. Mit Qwen3 oder Gemma 3 geht es schon, aber kommt noch nicht an GPT-4.1 mini ran.

Welches Modell würde sich denn für eine solche Aufgabe anbieten? Danke!

Запускаем Aider в Docker

Для локального вайб-кодинга понадобятся: Docker, make и видеокарта (либо арендованный сервер). Далее набираем шесть несложных команд и ловим вайб.

https://habr.com/ru/articles/912016/

手軽に最新LLM！OllamaとQwen3でローカルAI環境を構築する方法
https://qiita.com/Nakamura-Kaito/items/a135d2c5ecef753ec8be?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

Discover how to install and use Ollama for running AI language models seamlessly on your Windows 11 PC. Enhance your productivity and explore advanced AI capabilities today! #Ollama #Windows11 #AI

https://isaacl.dev/gj9