Est. MMXXVI No. 4 ◆ Archive ⌁ Free Models
LOCAL AI NEWS

Inference self-hosted on a DGX Spark · models served via LiteLLM

The Theragun Sense makes everyday recovery surprisingly easy — techcrunch.com
Nvidia’s AI edge now extends beyond its GPUs — techcrunch.com
Running a Chatbot on Your Own Computer — wired.com
Line 1: Chinese Automakers Follow Tesla's Bet That Robots Are the Next Big Profit Machine — techcrunch.com
President Trump Interrupted Nvidia's All-Hands Meeting to Call CEO Jensen Huang — wired.com
The Theragun Sense makes everyday recovery surprisingly easy — techcrunch.com
Nvidia’s AI edge now extends beyond its GPUs — techcrunch.com
Running a Chatbot on Your Own Computer — wired.com
Line 1: Chinese Automakers Follow Tesla's Bet That Robots Are the Next Big Profit Machine — techcrunch.com
President Trump Interrupted Nvidia's All-Hands Meeting to Call CEO Jensen Huang — wired.com
Editor's note — Light news day — several big items from the past week (GLM-5.3 open weights, Qwen3.8-Flash-Next, Open WebUI v0.11.1) were already covered, so this issue leans on fresh releases and funding news.

Tools & Runtimes

02

Bolnee-Chat: self-hosted RAG chatbots on your own LLM

Self-Hosted Stack

02

MCP Assist cuts Home Assistant voice LLM tokens by ~95%

Hardware & Edge

02

Qualcomm unveils Dragonwing QCS6490 for edge AI and industrial IoT

Free Models

24

Free means $0 per token today, not unlimited or unconditional: both platforms rate-limit free use and require a free API key, some OpenRouter free endpoints may train on your prompts unless you opt out in account settings, and the roster changes daily. Verified against each provider's own pricing data at generation time — but pricing changes without notice, so confirm it's still free on the provider's own page before you rely on it.

In Brief

03

Source: