frontpage.

The operational overhead of self-hosting model inference (vLLM setup, networking, scaling) is significant for small teams. A growing category of "MaaS" platforms abstracts this, so i've been testing hpc-ai.com for this purpose & its been working well. Built on Colossal-AI (~41k GitHub stars), & external endpoints with minimal config. Curious what HN thinks: right abstraction, or too much control lost?

Wisconsin governor says 'no' to age checks for porn

Show HN: An AI-powered knowledge base that thinks

Quansloth Using Google's Turboquant Breaks the "VRAM Wall" for Local LLMs

JSIR: A High-Level IR for JavaScript

Deterministic O(1) coordination via Recursive Field Invariants

MAIP – Machine Agent Identity Protocol

Show HN: Browser-based M68K interpreter supports MOVEM,data defs,and disp. addr

Show HN: Snill – Describe your business, get a complete internal system

We parsed 1,031 WA breach filings. Malware hides 5x longer than ransomware

Bridddge – Control LinkedIn Automation from Claude Code

An Arctic Road Trip Brings Vital Underground Networks into View

Sorting performance rabbit hole

Writing an LLM from scratch, part 32h – Interventions: full fat float32

The browser extension API setUninstallURL violates user privacy

Show HN: Niquests – A Requests fork, three years of catching Python's HTTP up

Blackholing My Email

How to Make a Complete Map of Every Thought You Think (2003) [pdf]

Every GPU That Mattered

Show HN: AI agents that learn from each other's mistakes

Running Out of Disk Space in Production

Your Agent's Context Window Is Not a Junk Drawer

Turn off your computer the only viable way: by flipping it off

Purple: Terminal SSH Client for macOS and Linux

Gemma4.java: Run Gemma 4 in pure Java (no Python, no JNI)

Ray – OpenSubtitles upcoming auto translating/subtitling video player

Building Slogbox

Austin VS Code hits v1, brings profiling MCP, live attach, and more

Show HN: Berroku – A fresh logic puzzle that's delightfully addictive

Ask HN: Gmail's missing paste-to-link feature

Ask HN

The case for Model-as-a-Service over self-managed inference