Show HN: Qwen3.8-27B API, 140 tok/s on one GPU inference.tiyuvta.ai 1 points by anotherCodder a day ago · 1 comment Reader PiP Save No comments yet.