I still can't believe I have this quality of AI on one DGX Spark.
NVIDIAAI Qwen 3.8 Flash Next NVFP4, on vLLM
Why is this 20-year-old kid's opinion so valuable that it's being propagated by every media outlet?
This has got to be the dumbest psyop I've ever seen. Even watching how simplistically retarded his analysis is makes him hard to watch, let alone take seriously. So some super
The benchmark was 2 websites, one LLM hosting platform and a travel platform.
The team was astonished at what Qwen 3.8 27B could accomplish, that some of them that are using Opus 5 even said they would use this model along side their workflows.
After 3.5 hours, the M3 Ultra finally one shotted Pac-Man!
Qwen 3.8 Flash Next. Pretty impressed. I might switch to local AI for 1 week.
You have reached the end of the archive
All of qwen38