FRAMEWIREIndonesiaUpdated Aug 23Live wire
0:00 / 0:00

My harness test, model : deepseek-v4-flash-vision-exp

1st one Deepseek harness 2nd one Hermes harness same prompt same model same thinking effort bijanbowen

Sooraj NarayanAug 231
0:00 / 0:00

China’s rapid AI advances are escalating the global tech race.

Former DHS Chief Privacy Officer Nuala O'Connor breaks down why protecting U.S. critical infrastructure is more urgent than ever ⬇️

PC MaticAug 23
0:00 / 0:00

Claude code for free (for real)

There is a repo with almost 48,000 stars that turns Claude Code into a client that you can use with free models. It's called free-claude-code. What it does: → You install a local proxy → Claude Code (terminal, VS Code or JetBrains) points to that proxy

Nix0nAug 2314
0:00 / 0:00

7.5 tok/s. That is what 512GB of DDR5 actually buys you on GLM-5.2, and it is the number that breaks my own advice from yesterday.

I said buy RAM before the next GPU. Then someone who owns the RAM row ran it. Threadripper, 512GB DDR4, RTX PRO 4500 with 32GB instead of the 96GB

BountyAug 2335