Show HN: Qwen3.8-27B API, 140 tok/s on one GPU(inference.tiyuvta.ai)
1 pointby anotherCodder5 hours ago1 comment