Quasar is an inference engine for Qwen3.6-27B-MTP / Qwen3.8-27B on a single RTX 5090, in the spirit of antirez/ds4
// readme
Quasar
A narrow inference engine for Qwen3.6-27B-MTP and Qwen3.8-27B-MTP (hybrid GDN+attention, trained-in MTP heads) and their fine-tunes on a single RTX 5090 (3090 and 4090/Ada also supported; Apple-silicon Metal backend for the q4s tier). One model family, one GPU, as fast as possible. In the spirit of antirez/ds4
Why this is interesting
-
Fastest of the four engines tested, on this harness — across two independent runs. Mean wall per SWE-bench instance, same 12 tasks, one harness, unchanged competitor binaries:
leg 2026-08-17 2026-08-19 q27 q5f 46.8 s 46.3 s q27 q4s 48.5 s 49.6 s llama.cpp 71.4 s 59.9 s vLLM 84.5 s 78.8 s ninfer NVFP4 96.8 s 113.8 s ninfer int8 327.4 s 266.2 s Two runs is what it takes to see the honest shape: q27 is the only leg that reproduced (46.8 -> 46.3 s), while every competitor moved 7-19% on identical binaries. So the ordering is stable but the margins are not — q27’s edge over llama.cpp reads 1.53x one day and 1.29x the next. Quoting either as the number overstates the precision.…