Look at these cool things our AI agent powered by Qwen 3.8 27B made in just a short while and without…
Look at these cool things our AI agent powered by Qwen 3.8 27B made in just a short while and without any feedback or input from a person so real world use will be much better if you send 1 or 2 extra messages to make improvements to the initial projects. First up is Hex-A-Gone!
My Qwen 3.8 Flash-Next Q4 runs on a 128 GB Strix Halo slowed with growing context across three harnesses.
That made me rethink the harness theory. Full follow-up:
Done with qwen 3.8 next flash using strata by coldniko .
You can check its stats in the presentation he mades
Every prompt you send to the cloud leaves the building.
Do you know where it goes? One team's agent sent a customer ticket to an outside model. Legal shut everything down at midnight. So for 7 days, my agents run on a Mac Studio on my desk. No cloud. Ollama + Qwen 3.8 27B.
You have reached the end of the archive
All of qwen38