Loading activity…
@0xsero
do you want to be free?
Technology
The company says the model improves benchmarks, task speed and output length over its predecessor.
AI
Artificial Analysis says Super VL uses about 60% fewer output tokens per task than its predecessor.
AI
A user says they're making a basic Electron version of the app, which they say will work everywhere.
AI
Underdog says Saluki is almost 7x smaller than its full-precision counterpart while retaining 96% of its benchmark performance.
AI
One post claims the ban followed fine-tuning and de-censoring an open-weight model for his agent harness.
AI
A user says they had local AI running on a brand-new machine in less than 20 minutes from boot.
AI
DHH says the collaboration aims to make local models work out of the box on Omarchy, with sponsorship through 0xSero’s company, Sybil Solutions.
AI
A user calls it an “unbelievably exciting time” and says local AI makes it even more so.
AI
A user links to an Omarchy local-AI plugin and says the setup needs no configuration.
AI
The user urges readers to sign a petition, framing ownership of computing resources and token generation as rights worth protecting.
AI
The creator says planned updates include support for multiple graphics processors and the ability to choose from a catalogue of AI models.
Technology
Critics link the Linux distro to far-right causes while DHH denies any political intent.
AI
The user reports an 86% GPQA score for GLM-5.3, saying most failures were timeouts rather than incorrect answers.
