Two Qwen3 models on one DGX Spark: the residency math
(news.ycombinator.com)
1.
2.
Running local models is good now
(news.ycombinator.com)
3.
4.
A 10 year old Xeon is all you need
(news.ycombinator.com)
5.
A 10 year old Xeon is all you need (for 26B-A4B MTP Drafters without GPU)
(news.ycombinator.com)
7.
Show HN: Forge – Guardrails take an 8B model from 53% to 99% on agentic tasks
(news.ycombinator.com)
8.
Running local models on an M4 with 24GB memory
(news.ycombinator.com)
9.
OpenWarp
(news.ycombinator.com)
10.
Stop Using Ollama
(news.ycombinator.com)
11.
April 2026 TLDR Setup for Ollama and Gemma 4 26B on a Mac mini
(news.ycombinator.com)
12.
Running local models on Macs gets faster with Ollama's MLX support
(arstechnica.com)
13.
Ollama is now powered by MLX on Apple Silicon in preview
(news.ycombinator.com)
14.
Show HN: Atomic – Self-hosted, semantically-connected personal knowledge base
(news.ycombinator.com)
15.
Show HN: Atomic – self-hosted, semantically-connected personal knowledge base
(news.ycombinator.com)
17.
18.
20.
175K+ publicly-exposed Ollama AI instances discovered
(news.ycombinator.com)
22.
Show HN: EuConform – Offline-first EU AI Act compliance tool (open source)
(news.ycombinator.com)
25.
New coding models and integrations
(news.ycombinator.com)
26.
Finding thousands of exposed Ollama instances using Shodan
(news.ycombinator.com)
28.
29.